YT YouTube2Text

YouTube2Text → Transcripts → AI and the Future of Work: Implications for Skills, Jobs, and Public Sector Readiness

AI and the Future of Work: Implications for Skills, Jobs, and Public Sector Readiness — Transcript

by Philippine Institute for Development Studies · 23,115 words · 3,615 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:02Good morning everyone and welcome to
0:04today's webinar organized by the
0:06Philippine Institute for Development
0:08Studies or PIDS Department of Labor and
0:11Employment Institute for Labor Studies
0:14and held in partnership with the
0:16socioeconomic research portal for the
0:18Philippines. SERP is a free open access
0:21library of policy research convened by
0:24PAS.
0:26In this age of AI where information can
0:28be misleading or unverified, SERP gives
0:31you trustworthy evidence-based studies
0:34from over 75 institutions. Complementing
0:37this, the Institute for Labor Studies as
0:40a policy research and advocacy arm of
0:42the Department of Labor and Employment
0:45contributes evidence-based research on
0:47labor and employment issues to help
0:49inform policy, strengthen programs, and
0:52support Filipino workers in a rapidly
0:55changing world of work. Our webinar
0:57today is titled AI and the Future of
1:00Work: Implications for Skills, Jobs, and
1:04Public Sector Readiness. My name is
1:06Marvin Kabangunai from the Department of
1:08Labor and Employment Institute for Labor
1:11Studies SRP Network member and I will be
1:14your host and moderator for this
1:16morning's discussion. We are very
1:18pleased to be joined today by
1:20participants from various sectors
1:23including representatives from national
1:25government agencies, local government
1:28units, academic institutions, the
1:30private sector, civil society
1:33organizations, and of course our
1:35partners from the media. Our discussion
1:38this morning is both timely and
1:41important. And uh as artificial
1:43intelligence continues to reshape
1:45economies around the world, it is also
1:48transforming how we work, the skills we
1:51need, and how governments deliver
1:53services. In the Philippine context,
1:56these changes raise important questions.
1:59Are our workers equipped for the future?
2:03How will jobs evolved or be displaced?
2:06and how ready is our public sector to
2:09adopt to AI technologies. This
2:12conversation also builds on our recent
2:14discussion on poverty and vulnerability.
2:17While many Filipinos are not classified
2:20as poor, a large share remain at risk of
2:23falling into poverty. As AI reshapes
2:26labor markets and it can either reduce
2:29this vulnerability or even widen
2:32inequalities if workers are left behind.
2:35Today we will share insights from recent
2:38studies from global trends to the
2:40Philippine context focusing on AI skills
2:44and the future of work and what these
2:46mean for inclusive and sustainable
2:48policy. We encourage everyone to
2:51actively participate in today's
2:53discussion. You may send your questions
2:56through Zoom Q&A box and we will address
2:58them during the open forum later. So
3:01before we begin, let us first
3:03acknowledge some of our distinguished
3:05speakers and guests joining us this
3:07morning. First of course we have the
3:10Institute for Labor Studies Executive
3:12Director Tri Janette Damo.
3:16We also have uh development academy of
3:18the Philippines acting president and
3:20chief executive officer Leoadadio
3:23Sebastian
3:26Center for Migrant Advocacy Executive
3:28Director Elen Sana.
3:31We also have Philippine Commission on
3:33Women Executive Director Narlene Santos
3:36Millier,
3:38Technical Education and Skills
3:40Development Authority Executive Director
3:43Charlene Justin Baste.
3:46Kapangan Manakaran Incorporated
3:49Executive Director Vivian Dabu.
3:53Always Trade Union Congress of the
3:55Philippines President Susanita Desorna.
3:58And lastly, we have Philippine Consulate
4:01General in San Francisco, Ambassador
4:04Rowena Pangilan Dakupil.
4:07To formally open our webinar today, it
4:09is my pleasure to introduce the
4:11president of the Philippine Institute
4:13for Development Studies. Please join me
4:16in welcoming Dr. Philip Arnold Tuano for
4:18his opening remarks.
4:22Salamat Marvin, distinguished
4:25colleagues, partners in policy research
4:27and friends from government, the academ
4:30private sector, civil society and
4:32development organizations.
4:34Good morning.
4:36Philippine Institute for Development
4:38Studies
4:41webinar
4:44AI and the future of work implication
4:46for skills, jobs and public sector
4:49readiness.
4:51This webinar is held in partnership with
4:52the PIS social economic research portal
4:55for the Philippines or SERB and our
4:59co-organizer the department of labor and
5:01employment institute for labor studies
5:03or do ILS.
5:05We would like to especially thank
5:07Professor Alexis Atinades for reaching
5:09out to PIS and expressing his interest
5:12in presenting his research work.
5:15This initiative has allowed us to
5:16convene this today's timely discussion
5:19and connect global insights with the
5:20Philippine policy context. We are
5:23pleased to have him with us later to
5:24share his insights on this important
5:26topic. Artificial intelligence or AI is
5:29already reshaping the way we work,
5:31learn, govern and interact. Around the
5:34world, AI adoption is accelerating at an
5:36unprecedented pace. According to the
5:39International Monetary Fund, nearly 40%
5:41of jobs globally could be affected by
5:43AI. In advanced economies, the share is
5:46even higher, reaching about 60% of all
5:49jobs. While some occupations may become
5:52more productive and efficient, others
5:53face the risk of displacement or
5:56significant transformation.
5:58Meanwhile, a recent report by the World
6:00Economic Forum estimates that by 2030
6:03around 170 million new jobs may be
6:06created globally due to technological
6:08change and new industries, but around 92
6:11million existing jobs may also be
6:13displaced. This figure remind us that
6:16the future of work is not simply about
6:17whether jobs will disappear. The other
6:20is about job how jobs will change, what
6:23new skills will be required and whether
6:25workers, institutions and governments
6:27are prepared for this transition.
6:30In just two week in in fact just two
6:33weeks ago we convened a discussion on
6:35poverty and vulnerability where we
6:37highlighted how closely labor outcomes
6:39are tied to poverty dynamics. We
6:42emphasize that beyond income levels, the
6:44quality, stability, and adaptability of
6:46work largely determine whether
6:48households remain resilient or fall into
6:50poverty during times of shock. As
6:53technological change accelerates, this
6:55relationship becomes even more critical.
6:58Workers who are unable to keep pace with
7:00evolving skill requirements risk left
7:03being behind. enforcing existing
7:06inequalities and vulnerabilities.
7:08This highlights the urgency of ensuring
7:10that our labor force is not only
7:12protected but also equipped to navigate
7:14and benefit from rapid technological
7:17developments.
7:22The country has a young dynamic labor
7:24force with over 50 million Filipinos
7:26participating in the labor market. At
7:28the same time, many of our workers
7:30remain concentrated in routine
7:32repetitive and administrative tasks that
7:33may v be vulnerable to automation.
7:37Studies have shown that occupations in
7:39clerical work, customer service,
7:42encoding, administrative support, and
7:44even some professional services are
7:46increasingly exposed to AIdriven tools.
7:49At the same time, we know that AI also
7:51creates opportunities. It can help
7:53improve productivity, expand access to
7:55services, reduce the cost of routine
7:58tasks, and generate demand for new roles
8:00in areas such as data analytics, digital
8:02services, cyber security, AI governance,
8:06and advanced technical skills.
8:08The challenge, therefore, is not simply
8:10to blindly adapt AI, but to ensure that
8:13adaptation is actually inclusive.
8:17We must ensure that workers are given
8:19opportunities to reskill and upskill. We
8:22must ensure that schools and training
8:23institutions are preparing learners for
8:25future jobs and must we must ensure that
8:28government institutions have the
8:29readiness infrastructure and governance
8:32frameworks need to harness AI
8:34productivity.
8:43It has the potential to improve service
8:45delivery, strengthen planning and
8:47support faster decision making and help
8:50governments respond more effectively to
8:52citizen needs.
8:54However, realizing these benefits
8:56require more than technology alone. It
8:59requires reliable digital
9:00infrastructure, skilled personnel,
9:03adequate investments, strong data
9:05governance, and clear ethical standards.
9:13This morning we will hear insights from
9:15three important studies. The first
9:17presentation is professor Alexis titled
9:20AI skills and the future of no work. It
9:23examines global trends in AI adoption
9:25labor markets and skills transformation
9:29using large scale international data.
9:32Building on this global insights the
9:34next study that will presented brings a
9:36discussion closer to home. The pad study
9:39how ready are loose for adoption by Pads
9:43uh researchers uh senior fellow Francis
9:46Marima, Christopher Ed Camp and Aliyah
9:49Mesa Salazar looks at the preparedness
9:51of local government units to prepare AI
9:53technologies and identifies the barriers
9:56that continue to limit their readiness.
9:59Finally, grounding these stud
10:01discussions at the level of the workers
10:02and industries themselves to study the
10:05impact of artificial intelligence on the
10:07labor market by Dole ILS senior labor
10:10and employment officer Chelsea Nicole
10:12Pineda explores how AI is shaping jobs,
10:16work experiences and skill requirements
10:18across key sectors in the country.
10:21Taken together, these studies reminds us
10:23that AI is not only a technological
10:25issue. It is a labor issue, governance
10:28issue, an education issue, and
10:29ultimately a development issue. As in
10:32many of our previous discussions,
10:34evidencebased policym will be essential.
10:38The decisions we make today on
10:39education, training, labor policy,
10:41digital infrastructure, and public
10:43sector modernization will shape whether
10:45AI will become a source of greater
10:47inequality or a tool for more inclusive
10:49or and future ready growth.
10:52We hope that these discussions
10:54contribute to a deeper understanding of
10:55the opportunities and challenges and may
10:58it help guide more responsive
10:59forwardlooking and inclusive policies
11:01for the Filipino workforce and the
11:03public sector.
11:06I now turn over to our moderator.
11:11Thank you very much uh sir Randy. uh as
11:14mentioned
11:17artificial intelligence or AI
11:21economy and even our societies
11:23worldwide. So may opportunities to lead
11:26increased productivity at new jobs uh
11:29may concerns like job displacement,
11:32skills gaps at unequal access of
11:35technology in the Philippines. While AI
11:38adoption is underway, gaps in skills,
11:41infrastructure, and institutional
11:42capacity suggest that we are not yet
11:45fully prepared for its widespread
11:47impact. Before we proceed, we would like
11:50to hear your initial insights on today's
11:52topic. As part of this, we have prepared
11:55a short call. So to participate, you may
11:58scan the QR code shown on your screen or
12:01uh click the link provided in the
12:03comment section. The question is, do you
12:06feel prepared to work with AI
12:09technologies in your current or future
12:11career?
12:13So
12:16to answer and access the poll question
12:19comment box again our question is do you
12:23feel prepared to work with AI
12:25technologies in your current or future
12:28career?
12:30So at the moment maximize allowable
12:33participants Zoom uh we're very happy we
12:37now have uh 14 participants and over 160
12:42viewers joining us via Facebook live.
12:44Thank you so much for joining. Uh the
12:47tech team is uh still receiving emails.
12:54So we encourage those who can't join on
12:56Zoom to join us on the Ps
13:00and Dole ILS Facebook pages.
13:04An soita talaga
13:07relevant and
13:09about AI and the future of work. Thank
13:12you so much for taking the time.
13:16Important conversation nto
13:19an so
13:22flash on the screen is our results of
13:25the pollag.
13:29So looking at the results
13:33hello perspectives
13:35participants which is actually a good
13:37thing.
13:42feel to work with AI technologies
13:48of awareness and readiness among our
13:51participants
13:54somewhat prepared and okay. It shows
13:57growing familiarity with these uh
13:59technologies
14:01and
14:03unsure or not yet confident and that's
14:06completely understandable actually.
14:10important initiatives like today's
14:13webinar uh creating spaces where we can
14:15learn together.
14:24Okay. So once again, thank you so much
14:27for your uh responses. Very insightful
14:30answers and emphasis on job security and
14:33skills which are very which are really
14:36at the center of discussions on
14:38artificial intelligence and in fact
14:41changes happening in real time. Jobs are
14:45evolving and some are at risk of being
14:47automated. New skills are becoming more
14:50in demand and uh organizations including
14:53government are being pushed to adapt
14:55more quickly. And this brings us to the
14:58central question of today's webinar. How
15:01is AI reshaping work skills and
15:04governance in the Philippines? And more
15:06importantly, who stands to gain and who
15:10might be left behind.
15:12Now we move on to our featured studies.
15:16Our first presentation is titled AI
15:19skills and the future of no work. The
15:22presentation examines how artificial
15:25intelligence is transforming labor
15:27markets globally from shifts in job
15:30roles and skills demand to its
15:32implications for inequality, youth
15:35employment and the future of work. This
15:38will be presented by Professor Alexis
15:41Antonades.
15:42Dr. Dr. Alexis Antonades is a professor
15:45and chair of international economics at
15:47Georgetown University in Qatar and a
15:50fullbrite scholar with a PhD from
15:53Colombia University. His research
15:56focuses on artificial intelligence
15:58skills and the future of work with prior
16:01experience at the Federal Reserve Bank
16:04of New York and Princeton University.
16:06Without further ado, let us welcome
16:09Professor Anton Yades.
16:14Thank you so much Marvin. Uh thank you
16:18uh Philillip for the invitation and
16:20thanks to the Philip uh Philippines
16:23Institute for Development Studies. I am
16:26really excited to be here and talk about
16:29AI and how it affects labor markets.
16:33I've been to the Philippines many times
16:36and for the reasons why um you will see
16:39in my slides. It's it's very important
16:42that we have this discussion. So what I
16:44would like to do over the next 25
16:47minutes is give you and the viewers a
16:52lot of information, a lot of data to
16:55start making some connections based on
16:58everyone's own experiences.
17:01I hope you will find that interesting
17:03and then we will be able to have a very
17:05interesting discussion in the open forum
17:08toward the end and uh to give you the
17:12let me share my slides.
17:16So I'll be wearing four hats and you
17:19will see what this means. Okay. Why is
17:22this such an important topic to discuss
17:25now?
17:27Well, what is very important is to
17:31realize that the rate at which new
17:34technology is being created has
17:36accelerated.
17:38So throughout humankind, we had new
17:41innovations, new technology. But the
17:43rate at which we we hear about new
17:46things, clothes is releasing uh clo
17:49design, clo every day we have something
17:52new that changes everything for us.
17:54That's uh unprecedented. The other thing
17:57is the rate at which technology has been
18:00adopted and is being in impactful has
18:04accelerated. So in the past we had the
18:07computer revolution but it took decades
18:10because before we all we saw the impact
18:14uh in the economy we had the dotcom the
18:16internet revolution. It took decades
18:19maybe less decades to see the impact.
18:22Now with AI, we have a new tool launched
18:25today. Tomorrow we have the impact.
18:28Tomorrow we'll see companies
18:30uh experiencing and dropping their
18:32valuation because that tool renders them
18:35obsolid.
18:37And and and these two facts, the rate at
18:39which these new technologies are being
18:42created and adapted causes fundamental
18:45shifts in labor markets where we work in
18:47our welfare, in business, in humanity.
18:50And that's why we need to study that.
18:53So I will start first with my academic
18:56hat and I will I will put this in
18:58context by giving you some very
19:00important data that you are all aware of
19:03but it's good to to summarize. So the
19:05first thing we know is that inequality
19:07is rising and that is not a new
19:10phenomenon right and it's important to
19:12start with that because when we want to
19:14talk about the future of AI well is it
19:17going to decrease in equality or is it
19:19going to increase what is our baseline
19:21so today I'm going to show you our
19:23baseline so what you can see on this
19:25graph and I will not have much time to
19:27go through all the details is that for
19:30the very rich the dark blue their wages
19:33their income has been growing much
19:35faster than any other income bracket. So
19:39the poorer you are, the lower income
19:42growth you experience. So we have a lot
19:44of inequality rising. Will that continue
19:46or will that be reversed? The second
19:48thing of course we care about income
19:51disparities because higher income is
19:55linked to higher life expectancy,
19:58lower child mortality. So as you move to
20:01the right is higher income, lower
20:03maternal death, more life life
20:06satisfaction, more years of schooling,
20:09less hours work. So we care about income
20:13and we know that up until now
20:17the people with the highest degree would
20:21get on average higher salary so with
20:23postgraduate than people with bachelor
20:26and people with bachelor would get
20:28higher salaries than people with some
20:30college degree and and better than high
20:34school dropouts.
20:37So decrease did matter and that may
20:39start to change.
20:41The other thing we know from the
20:44literature from many studies is that on
20:46average women earn less than men.
20:50In the beginning of the 80s they earned
20:53about 38% less.
20:55Over time that close to about 22%.
20:59We call this the gender wage gap. And it
21:02could be that women are willing to
21:05accept the lower salary because they
21:08care more about the benefits, the
21:10ability to go home, take care of the
21:12kids, have some flexibility. But we know
21:14that the wage gap
21:16has decreased, which is great news, but
21:19it hasn't moved for like three decades.
21:22So how will AI change that? Is it going
21:25to continue to reduce the wage gap
21:28or is it going to increase it? Right? So
21:30that's our benchmark. Now why do women
21:34earn less? One may speculate that women
21:38don't have the same opportunities are
21:41meant to get education and that explains
21:44why. So on average men have higher
21:47degrees
21:48than women. That's not the case. So even
21:51if we split the sample by age by degree
21:56brackets.
21:57So every column is a degree by bracket
22:00like bachelors.
22:02Men blue on average earn more than women
22:07after we control for education. So it's
22:08not really education.
22:10And of course in STEM, science,
22:12technology, engineering, and math,
22:15there's a higher representation
22:17in in in in advanced economies of men,
22:21not women. And that we care a lot
22:24because a lot of the newly created jobs
22:26are in STEM and they give higher wages.
22:29So if only men pursue STEM, they will
22:32get the highest paying jobs and the wage
22:34gap is going to increase. But um and
22:37yeah, let me skip now. I'll come to
22:39that. So why is it that women on average
22:41earn less?
22:44What you see here is on the left color
22:47coding the inequality. So
22:49[clears throat] many colors means a lot
22:51of inequality.
22:53So if we adjust for education
22:55differences,
22:56well, we don't see a difference. There's
22:58still a lot of inequality.
23:01However, if we adjust for occupational
23:03choice, we see that a lot of the yellow
23:07orange colors go away. So inequality
23:10between men and women in wages drops. So
23:13what this means and that's very
23:16important we'll carry that with us as we
23:18think about the impact of AI is
23:20occupational choice matters. It could be
23:24that women decided to become teachers
23:26because they love the profession and men
23:29with exact same education decided to go
23:32and work for a bank.
23:35Now it could be that the wages in the
23:37bank are moving much faster than in
23:40schools. So what we see is that the wage
23:42gap between teachers and bankers is
23:45increasing and because we have a lot of
23:47men bankers, male bankers, the wage gap
23:50is rising. So occupational choice is
23:53very very important and we will come
23:55back to this a couple of times. So keep
23:57that with you. Now we have the superstar
24:00firms. We don't have time to discuss
24:02that. And I will conclude that in in in
24:06urban cities wages are rising faster
24:10than in rural cities
24:12this is the density. So basically this
24:15is very low density that's in the middle
24:17of nowhere. It doesn't matter if you you
24:20know between 2015 and 1980 the wage
24:24growth is the same. But if you go in a
24:28high density city, wages in 2015 rose
24:32much faster than other zones. What is
24:34driving that? Business services. In the
24:37middle of nowhere, you don't have
24:39services.
24:41But in the in the big cities, you have a
24:44lot of positions in business services
24:47and the demand has gone up and that has
24:50resulted in higher wages. And that's
24:52relevant to the Philippines because the
24:54Philippines exports a lot of services
24:57like call centers, right? So we will see
25:00how AI is going to affect that.
25:03All right. In the second part, I put the
25:06hat of a researcher. So I will tell you
25:08a little bit about my work. In fact, I
25:10will tell you more about the data I use
25:12and then I'll show you some good
25:14findings.
25:16So um so how do we make sense of AI?
25:20Well, the first thing I've done with my
25:23colleagues, I got data on vacancies
25:26posted online. That is if a company or a
25:32government agency would like to hire
25:34people, they put a job post and that's
25:39online. So if I get that job post and I
25:42read it, I get some information on how
25:46the demand for different skills and
25:48different occupations is changing in the
25:51labor market. So I collected data on 1.5
25:56billion vacancies from the US and China
25:59between
26:012014 and 2022 23 and I'm continuing to
26:05collect more. So I'll show you results
26:08from the US. I have some China but no
26:10time to show you. So what do the the
26:14data look like? I can go into my Excel
26:16sheet and I can see all the jobs posted
26:19in the US in 2014
26:22and there are tens of millions and I can
26:24say how many of those jobs mention
26:27plumbing
26:29and I see different skills. Uh there
26:32were 605 jobs for plumbing design
26:36for mechanical electrical and plumbing
26:39design 1500 commercial 2000. Right? So I
26:44can I can see the different types of
26:47plumbing jobs not just plumbing which is
26:48very interesting. I can do the same
26:51thing 5 years later
26:54and I can see which skill is growing
26:57more in demand.
27:00[clears throat]
27:01I think someone has to mute their
27:03microphone. So I can see that for
27:05plumbing design demand increase 3.3
27:09times
27:11or 2.3 times more than what it was
27:14before. So that allows us to see the
27:16change. This is there's no no analysis.
27:19This is just counting before and after.
27:22So if you want to advise you know your
27:24kids what type of plumbing to study,
27:26that can be very useful. For example, I
27:29can look at massage and I can see the
27:31different types of massage companies
27:34were hiring for and I can see that while
27:38sport massage wasn't that big in 2014,
27:42it grew the highest in demand. Okay,
27:45that's very very interesting.
27:47I can look at other things. 3D printing
27:51grew this is 5.1 times. This is one plus
27:57this number. That's a lot. Amazon Web
27:59Services 7.3
28:02opioid. How many jobs mention in the
28:06description? Something about opioid less
28:09than a thousand. Just 5 years later, we
28:12have nine times more because of the
28:14opioid epidemic in the US. Bitcoin is
28:17growing. Major depression is very key.
28:22fintech, knowledge of marijuana,
28:24blockchain because blockchain didn't
28:26exist, right? So, this is me searching
28:29on my Excel sheet for different degrees.
28:34What if I what if I want to do the
28:36following? What if I say let me not
28:39search
28:41data please tell me what are the
28:43keywords what are the skills and I have
28:4616,000 skills that grew the most
28:50between skills that existed in 2014 and
28:53in 2022
28:56so show me that
28:59and what do I see it's words I don't
29:02recognize
29:05kypernet
29:06blockchain
29:08Salesforce SAPS for HANA
29:12SAP S4 HANA
29:15uh here because it has a space here it
29:17doesn't have
29:20so everything is about technology
29:24these are skills that existed so in 2014
29:28you had jobs in Salesforce
29:31but in 2019 you had way more jobs in
29:34Salesforce so the bigger the font sizes
29:37the more they increase.
29:40So it's all technology and we call this
29:43a word cloud. All right. If I say hm
29:47what if I I look at
29:51new new keywords new skills that
29:53appeared in 2019 that didn't exist 5
29:57years earlier.
29:59again technology tensorflow dev sec ops
30:03blue prism blockchain technologies for
30:06internet of things deep learning
30:09jupitter notebook
30:13um gdppl that's a new requirement by the
30:16European Union all right so what what if
30:19I ask okay that is cool can you tell me
30:23if there are any skills that were
30:24important in 2014 and by 2019 they were
30:28on the way out
30:32you know jobs requiring experience in a
30:35team transaction smell and taste I have
30:38no idea what it was it was before COVID
30:41but it's maybe a clinical trial in 2014
30:45that was hiring lots of people in that
30:47and then you know you don't have the
30:48trial I medical evacuation
30:52uh satellite placement assessment
30:57eBay enterprise
30:59BlackBerry Enterprise web practices.
31:02So these are the even corporate law like
31:05this very generic.
31:08All right. So now we start to get
31:11something from this data. Now I'm just
31:13looking at the data. I love to spend
31:15time just looking at the data. Not
31:18trying to say a story but trying to see
31:20what story the data is telling me. And
31:23um
31:26and one another thing we can do we can
31:28do the exact same exercise and that's
31:30getting us closer to our message by by
31:35occupational group by business
31:38operations. What are the skills that
31:42were, you know, grew the most in in
31:45business operation between 2014 and 2019
31:49and the ones that disappeared?
31:52What about engineers? What about
31:54lawyers? And now you start to see a
31:57picture.
31:59Regardless of the occupational group,
32:02the skills that are becoming important
32:04in terms of growth are very similar. But
32:08the skills that are disappearing were
32:10very specific to each occupation. So
32:12that would be the first
32:15uh in insight I'm going to give you from
32:17my research. By looking at 1.5 billion
32:21vacancies,
32:23my team and I were able to see that
32:26occupations, think of occupations as
32:28islands, they start to come closer to
32:30each other. So there isn't as much
32:34difference between teachers and bankers
32:36now as there was 10 years ago or
32:40engineers
32:41and economist. And the reason for that
32:45is that both engineers and economists
32:48need to know how to use Python or data
32:52analysis. Data analysis is very
32:53important for both. So they start to
32:56look more homogeneous.
32:59So as time goes by, there are some
33:01skills occupations no longer need and
33:04they drop them. And these skills skills
33:06are very different for from one
33:08occupation to the other. If you are an
33:10architect, you need to use um auto cut
33:15just for architects. As time goes by,
33:17there are new skills that are becoming
33:19important. Architects may need to know
33:22data analysis also, data visualization,
33:24but economists need to know that. In
33:27mining they need to know that in
33:29marketing they need to know that in real
33:31estate in finance so data analysis
33:34becomes important for all the skills for
33:37all the groups so groups start to look
33:39very similar now that's how I spend my
33:42time that's my academic hat but there is
33:45a limit to how much I can tell with my
33:48academic hat because I I rely on on jobs
33:53posted to see what's coming ahead
33:56But it means companies are already
33:58posting. What if I want to see the
34:00future before the future comes? So here
34:03I'm wearing my VC fund.
34:06So I have a I have some friends they
34:08have a VC fund and it invests in Silicon
34:12Valley
34:14accelerators in Y Combinator.
34:17So they are able to see the best
34:19startups that come in before anybody
34:21else and 95% of their investment 96 are
34:25in AI. So they make investments and that
34:28allows me to see what will be available
34:30in 2 three years from now and that
34:33allows me to combine for example I'll
34:35mention only one I'll mention two solve
34:40intelligence is a company that uses AI
34:43to uh to solve the problem of patent
34:47filing. It does all the research. It
34:49completes all the documents using AI for
34:53different jurisdictions in different
34:55languages for different regulatory
34:58environments and it reduce the time
35:00taken to work on patents by law firms
35:0480%.
35:06So law firms don't need a lot of
35:09like entry- levelvel lawyers because
35:12this so intelligence can do most of the
35:15work and one senior lawyer can just stay
35:17there and just sign and check it. They
35:19even launch a product that if you are an
35:23individual and you are thinking of
35:24patenting something and you don't know
35:26how to start, you subscribe to this
35:30and to this service and it will do all
35:31the work for you. It will prepare the
35:33documents, everything. it will advise
35:35you and all you have to do is take it to
35:37a lawyer and get it signed. So it really
35:39changed the business for patent lawyers.
35:44So you will need significantly less
35:46patent lawyers moving forward than you
35:48need now. And so far they went for
35:51series A. So they are in the early
35:52stages and they already have contracts
35:54with the top 400 law firms in the US
35:58that do patent.
36:00Open Call is one of the five AI
36:03companies they invested that automates
36:06call centers and that's very important
36:09for the Philippines and and this is not
36:11the early versions where you get so
36:13upset because the voice doesn't
36:15understand what you are saying. Now with
36:18AI call centers can use AI to answer
36:22calls because they can switch languages.
36:24They can retrieve information in no
36:27time.
36:28Scale will not be a problem. You don't
36:30have to wait 10 minutes in line for the
36:32next and that will in that vertical. So
36:36if so intelligence will kill 80% of the
36:39jobs in patterns
36:41in the legal profession, call centers AI
36:45will kill 97%.
36:48And so on so it's really so that allows
36:51me to see to see more.
36:55Let me take this. I'll mention Domos.
37:00Domo is a Californian-based startup that
37:03automates call centers for collection.
37:06So, for example, if you got a bank loan
37:09and you are not paying, the bank will
37:12give you a call
37:14and will ask you, you know, why you're
37:16not paying and please make your
37:17payments. Uh you got a car loan and you
37:20are not paying. You have a credit card.
37:22you get these calls. Domu automates
37:25these calls using AI
37:28and what they found they did it in
37:30California that's where they started
37:33their results when they have the AI call
37:36people is much higher but then when they
37:38have they have human beings calling
37:41other human beings in terms of is was
37:45the call successful do they start making
37:47the payments and you get more success
37:49with AI because AI doesn't lose the
37:52temper they understand uh the other
37:55person so they can change their tone. Uh
37:58if they hear the other person with an
38:00accent like mine, they can switch the
38:02language and they have better results
38:04and so on. So that so what I'm trying to
38:07do is take my my analysis from my
38:11research data and my research and
38:13analyzing those new startups that will
38:17become some of them few of them billion
38:19dollar companies to understand which
38:21vertical in the labor markets are going
38:24to be destroyed by them unfortunately.
38:30All right. So if I look at that
38:33depending on the vertical you talk about
38:3510 to 97% of job loss across verticals
38:41and um and Philip talked about upscaling
38:45and rescaling and that is so important
38:48because things are changing very very
38:51fast
38:52and we don't have answers. One of the
38:56main lessons I try to pass to people is
39:00we will be lucky if we even know what
39:02are the right questions to ask. So
39:04having this type of conversation is so
39:07important, not because we have answers,
39:09but because it brings people that are
39:12thinking about this with people working
39:14in different sectors and we all exchange
39:16information and we try to contribute
39:19from our own perspective because you see
39:22things differently if you work in a call
39:24center, if you work in the shipyard, if
39:28you work in a bank, if you work as a
39:30secretary, if you work as a professor,
39:32but we all need to come together. All
39:35right. So, in the last five minutes, I'm
39:38conscious of time. What I want to do, I
39:41want to wear my philosopher hat. So, if
39:43I were to combine all the things I see,
39:48what are some of the lessons I take?
39:51Lesson number one, skills are becoming
39:56more important, not degrees.
40:00So the boundaries between degrees is not
40:03that important anymore. Before if you
40:06were studying banking or finance was
40:09different than engineering than
40:11economics.
40:13It's still different now but not that
40:14important because it's what classes did
40:17you take? Did you take data
40:19visualization? Did you take uh
40:20programming? Did you It's very important
40:22in every occupation. So what we find in
40:25the labor markets especially in advanced
40:28economies we see a switch from degree
40:30based hiring to skill-based hiring
40:34and we have the what's called emerging
40:37degree reset. Many companies like the
40:40ones you see on the right very big
40:42companies they drop the requirement that
40:44you have a degree when you apply to
40:47those companies. So they will take you
40:48even if you don't have degree. They
40:50don't seem to care about degrees because
40:52they think it's skills that matter.
40:56So that means for us as educators and
40:59for us as policy makers
41:02and even as parents, we must prioritize
41:05skill-based training. That doesn't mean
41:09of course we shouldn't send our kids to
41:12the university. That's very important
41:13because soft skills are becoming more
41:15and more important. But we also need to
41:17understand that skill-based training is
41:20important
41:24and and and and these skills by the way
41:26are not necessarily the ones we teach
41:28students at the university but they are
41:31skills they can learn online in very
41:34short time. So the other thing we found
41:37with our research is that occupations
41:39are converging. So the distance between
41:42occupation is coming closer and yes
41:46there was no measure of distance. What
41:48does it mean the distance between
41:49teachers and engineers? We built one we
41:52measured it and we showed that over time
41:54occupations are coming closer together
41:56and because they do that wages between
41:59occupations on average are converging
42:02and mobility is rising. So it's easier
42:05for people to jump from one occupation
42:07to the other. And that's the opposite of
42:09what all the theories were predicting.
42:12But we also found we used data on 5,000
42:15universities in the US for a 10-year
42:18period uh with about 3.5 to 4 million
42:22degrees awarded every year. So we had
42:24data on every degree, every person who
42:25got a degree in the US. We found that AI
42:28is pulling more students into STEM
42:30education,
42:32but when we look at the gender
42:33composition, it's pulling more men than
42:36women into STEM. And that's a little
42:38that's alarming because we want more
42:41women to go into STEM.
42:47We know there is a no the the next thing
42:51is
42:52that the the labor market is changing in
42:56what I think is three fundamental ways.
42:59The first one we see a mismatch between
43:01degrees and skills. Many companies
43:04require skills that we don't teach
43:07students at the university.
43:11And many students who graduate and they
43:13take surveys, they report that they
43:15don't feel they're well prepared for the
43:18labor force. So on both sides and in
43:21fact if you look at the responses from
43:24students studying computer science they
43:27will be the ones that are more vocal
43:29about not being well prepared which is
43:32the opposite of what we expect. The
43:34second
43:36finding is that there will be a big drop
43:39in demand for jobs because many of the
43:42jobs and many of the tasks we do within
43:45a job will be replaced by AI.
43:49I am building a startup from my laptop
43:53using cloth without the need to hire
43:56engineers. It's incredible. I have an
43:58idea. I'm very creative. Now there is no
44:00technical back barrier. I can use AI. So
44:03there will be a huge drop. And the other
44:06thing that is happening is in the
44:09example I gave with lawyers, you still
44:11need the senior lawyer, but you don't
44:13need to hire a lot of analysts. In
44:15banking, in in in Wall Street, in
44:18finance, you don't need to hire a lot of
44:20analysts. Uh for me as professor, I used
44:24to hire a lot of research assistants to
44:26collect data, uh to make tables, make
44:29some nice graphs, read the literature,
44:32summarize it for me. I can just do it
44:35with AI. So, I reduce the number of
44:38research positions I'm hiring. I'm still
44:41willing to hire somebody if I need
44:44something more technical. And that
44:46presents a third challenge.
44:48How am I going to find people that have
44:50more technical experience if we don't
44:53give them the opportunity to start from
44:55the bottom and move the way up? Learn.
45:00How am I going to end up with senior
45:02lawyers with experience 10 years from
45:04now if I'm not hiring a lot of
45:05entry-level lawyers? How am I going to
45:09end up with very experienced
45:12finance professionals if I don't get
45:14them analysts? So, what is happening? AI
45:17is not just limiting the number of jobs.
45:20It has disrupted this pipeline where you
45:22start from low. You know nothing. That's
45:24how I started trying to get internships
45:26and I knew nothing. But you learn and
45:28you gain experience and you start moving
45:30up. So now we have fewer jobs and the
45:34hurdle we need to jump is much much
45:37higher in order to get that job.
45:42Unfortunately, the benefits will go into
45:44specific countries and I'm finishing in
45:46two minutes. But the
45:50but the cost will go in other countries
45:52such as the the Philippines for example
45:54because of the impact on uh export
45:57services.
45:59And as the as the impact of AI on the
46:02labor market increases, some countries
46:04may be better positioned to support the
46:06labor force than others. And that's one
46:10challenge.
46:11And where does this lead us? I think
46:15because there will be fewer things to
46:17do. I see us moving into three to four
46:19day work weeks. I think governments will
46:21have to pay universal basic income
46:24because people will be doing less. But
46:26not all governments will be available to
46:28do it and that will become a problem and
46:31uh that will create a lot of inequality,
46:34a lot of instability, a lot of tensions
46:35in society
46:38and we will end up spending a higher
46:40share of our lives online for nonwork
46:42rellated activities and you see that
46:44from our kids. They're always in front
46:46of the screens not being you know very
46:49interested in reading books or studying.
46:52They just consume things online.
46:55And finally, I will conclude with this
46:56and I apologize for taking two extra
46:58minutes, but what we see is a generation
47:01of unhappy people. We see that uh people
47:05feel unhappy. They feel they've been
47:08left out and that will increase.
47:10Frustration, depression will be on the
47:12rise. We will we will have a feeling of
47:15relative economic decline even if the
47:17numbers look great because we compare
47:19ourselves with other people we see
47:21online that portray a very different
47:22image.
47:24And that leads to political
47:26polarization, populist voting,
47:29anti-global attitudes. We already see
47:31that. And of course that can have very
47:34adverse uh adverse effects. So I end up
47:37with this picture in 2015, 2005, 2010.
47:42They they have this survey where they
47:44ask people how happy are you? And
47:47people, you know, at 20 they are very
47:50happy. They're about to start college,
47:52you know.
47:53As they move through college, happiness
47:55goes down. But they are very happy.
47:58People in the 30s and 40s and 50s are
48:01not happy. They start family. They have
48:03kids. They they have so many expenses.
48:06They struggle to make ends meet. Around
48:0960, quality of life improves because you
48:13no longer have kids to take care. Maybe
48:15you have grandchildren. They did the
48:17same thing in 2022.
48:22this part, the ones that supposed to be
48:25the happiest
48:27are so unhappy now. And it doesn't mean
48:30that as age goes by, they become
48:33happier. No, this means that I ask
48:36somebody who's 40 years old now and they
48:39are happy. It doesn't mean that this 20
48:41year old when I ask them in 20 years
48:43from now, he or she will be up here.
48:45This is the previous generation.
48:48Things are changing. So with this in
48:51mind, I you know this is my email. You
48:54can find me on LinkedIn and you can feel
48:56free to connect with me. Thank you so
48:58much. Sorry for taking more time. We'll
49:01talk later in the photo.
49:04>> Thank you very much, Professor Alexis,
49:07for that insightful presentation. Uh one
49:10key takeaway is that as technology
49:13accelerates and reshapes jobs, we are
49:15seeing both new opportunities and
49:18growing inequalities. So this really
49:20underscores the urgent need for the
49:22governments, the institutions and
49:24workers to invest in continuous
49:26upskilling uh inclusive policies and
49:29human- centered transitions so that no
49:32one is left behind in the future of
49:34work. Once again thank you so much
49:37professor Alexis for your presentation.
49:39So first presentation po and
49:43on the future of work uh especially with
49:46the emergence of AI. Professor Alexis
49:49presentation will be posted ps website
49:53after this webinar. We now move to our
49:56second presentation and let's zoom in
49:59our lens to a more local context. Our
50:02next presentation features the PIDS
50:05study titled How Ready Are LGUs for AI
50:09Adoption? This study is authored by PIDS
50:13senior research fellow Francis Mark
50:15Kimba, supervising research specialist
50:18Christopher Ed Kabu Verde and former
50:21research analyst Aliya May Salazar. The
50:24study examines the preparedness of local
50:27government units in adopting artificial
50:29intelligence technologies.
50:31Findings suggest that while AI presents
50:34opportunities to improve public service
50:36delivery, most LGUs exhibit low to
50:39moderate readiness due to gaps in
50:41digital infrastructure, limited
50:44technical capacity and insufficient
50:46investments in ICT and skills. This will
50:50be presented by Dr. Francis Mark Kimba
50:53who is also the director of the
50:55Philippine Apex Study Center Network of
50:57PIDS.
50:59Dr. Kimba has worked on agriculture,
51:02trade and rural development and
51:04currently focuses on innovation among
51:06local firms. He holds a master's degree
51:09from the International University of
51:11Japan and both his MA and PhD in
51:15development economics from the National
51:17Graduate Institute for Policy Studies in
51:20Tokyo. Let us now welcome Dr. Kimba.
51:25>> Hi uh good good morning everyone.
51:29I will be presenting our paper on how
51:32ready are LG used for AI adoption as
51:34mentioned earlier. Um can you go to the
51:38next slot?
51:49>> So AI has emerged as one of the most
51:51transformative technologies in recent
51:54years. So in the Philippines,
51:57the AI sector is projected to grow from
51:59$772 million in 2024 to $3.4 billion in
52:052030. And this is uh from ITA in 2024.
52:09This is uh roughly an annual growth rate
52:11of 29%.
52:15Uh yet AI policy adoption in the Phil in
52:18ASEAN has lagged significantly behind uh
52:22OECD countries and within the
52:25Philippines, internet access also a
52:27concern. So um it remains concentrated
52:30in um the main island in Luzon with
52:33limited penetration rates in uh Visayas
52:36and Mindanao.
52:38So these gaps uh make it urgent to ask
52:41uh because of the the the potential but
52:44also the challenges being faced by AI.
52:47We we are asking uh our are our LGUs our
52:52local government units actually ready to
52:55adopt uh uh artificial intelligence.
53:01Uh in the next slide we want to frame
53:03the this
53:07uh okay um can you move to the next
53:09slide please?
53:12Okay. So,
53:16I'm sorry, pre previous slide, please.
53:18Sorry. Okay. So, we want to frame the
53:20discussion focusing on generative AI
53:23that can transform local governments
53:25across uh three tiers. So uh tier one uh
53:30where most uh applications actually sit
53:33generative AI functions as a digital
53:35assistant automating tasks like uh
53:38meeting summaries drafting job
53:39descriptions and others. Then tier two
53:42improves communication with residents
53:44through AI chat box chat bots uh
53:47automatic translations and improve
53:49access to public services. And then tier
53:51three tackles large scale challenges
53:54such as processing public um benefits
53:56applications um and others.
54:03>> Okay. Uh okay. So can we go to the next?
54:06Okay. So given this context the paper
54:08asks three uh main questions. First what
54:11is the current state of digital
54:12transformation technical capacity human
54:15resources in LGUs that would support AI
54:18adoption?
54:20Uh second uh what are the key barriers
54:22uh LGUs face uh in implementing AI
54:25solutions? And then the third uh how do
54:28organizational and reg regulatory
54:30frameworks uh at the local level enable
54:34or constrain the adoption of AI?
54:38So before we uh next slide please before
54:39we proceed to the findings um I think it
54:42is important to ground this to uh
54:45applications that we can really see or
54:47feel or real world applications. So it's
54:50um we look at uh cases in other
54:52countries um about uh the use of AI. So
54:56first is an international case the
54:58Singapore's um one service chatbot which
55:02allows residents to report incidents to
55:04provide feedback directly through
55:06WhatsApp and Telegram. So without
55:08needing to download a separate
55:10application.
55:12So this is actually a practical example
55:14of a tier 2 AI improving access to
55:17public services.
55:19Uh next slide please. So the chatbots uh
55:23VICA no so virtual intelligent chat
55:27assistant um automatically classifies
55:30the feedback and routes it to the
55:32appropriate government agency including
55:35photos, video attachments and others. So
55:37this is a significant improvement over
55:39traditional chat bots that force users
55:41to into predefined categories which
55:44often uh do not match their concern. And
55:47I know a lot of us have actually
55:49experienced this no using chat bots that
55:51somehow feel frustrating because they're
55:53trying to classify you into segments or
55:56areas that are are not really what you
55:59you are feeling know your concern.
56:03So the second uh example in uh Buenas
56:06Iris in Argentina more developing
56:10country than compared to Singapore. So
56:13which introduced botti bi no so a
56:16chatbot similar to in concept to
56:18Singapore's one service is designed to
56:20assist residents with city service city
56:23service inquiries and feedback.
56:26Uh in the next slide um BI went further
56:30than just inquiries. So when um not just
56:32inquiries it also processes government
56:34paperworks such as national ID
56:37applications, driver's licenses requests
56:39and others. So the screenshot on your
56:41screen shows what it looks like on the
56:43Buenos Iris uh city government uh
56:46website.
56:48So you can see there are a lot of the h
56:52um uh available options for BOTI for
56:55BOI. So critically um next slide please.
56:58By was embedded directly into WhatsApp.
57:01So similar to what Singapore did not so
57:03the most widely used uh messaging app in
57:06Argentina which dramatically lowered the
57:09barrier for residents to access uh city
57:11services because you can now um directly
57:14use it or and um using the platform that
57:19is most available to the residents in in
57:21the city.
57:23Um next slide please. During COVID 19
57:26the uh the pandemic body was even used
57:28to detect infection by analyzing cough
57:31audio clips via WhatsApp with an
57:33accuracy rate of around 88%.
57:36However, Botti also experienced system
57:38breakdowns during peak usage and u
57:41especially when during COVID no when
57:44they had to book uh appointments for
57:47vaccination. So that's actually um
57:50resulted to a breakdown. So reminding us
57:52that even mature AI tools have
57:54scalability limitations and this is
57:57actually important LG for LGUs an
58:00important lesson for LGUs and even for
58:02the Philippines considering to implement
58:04um such applications.
58:07Okay. So
58:10there there are applications present but
58:13let us see now for our local context how
58:16ready are our LGUs.
58:19So in the next slide we mentioned very
58:22quickly our methodology know. So for our
58:24methodology we looked at our readiness
58:27index. We constructed a readiness index
58:30um using a small data set the 2023 DICT
58:34satellite survey data set and the index
58:37synthesizes six dimensions as you can
58:39see here uh in this slide into
58:41comprehensive scores ranging from 0 to
58:44100 or or 0 to one. So um enabling uh
58:50comparisons uh about uh different uh
58:53LGUs. In the next slide we present the
58:57six pillars. So digital infrastructure
59:00we give that a weight of 20%. Um
59:04covering connectivity, network quality
59:06and ICT equipment. Human capital and
59:09skills are also 20% and this covers the
59:12workforce readiness in the LGU and
59:14digital literacy.
59:16Innovation ecosystem also has 20%
59:19measuring technology adoption
59:21digitalization progress
59:24and then governance and policy at 15%
59:28covering strategic planning and
59:30institutional capacity data readiness is
59:3310% measuring e-services data management
59:36and economic foundations at 15%
59:39measuring the fiscal capacity for
59:41sustained AI investment.
59:44So the specific subindicators for each
59:47pillar are are actually displayed uh are
59:51actually discussed in in in in the
59:53paper. In the next slide we show the
59:56coverage. No. So uh the survey covered
1:00:0035 uh municipalities. So uh the
1:00:04population of interest is presented here
1:00:06compared to the sample. There are 296
1:00:09also municipalities that um were covered
1:00:12and then 17 provinces out and this is
1:00:16out of a total of 1,695
1:00:19LGUs we have around 347
1:00:22who participated in the survey.
1:00:26In the next slide the let me just
1:00:28provide you a summary of of our results.
1:00:31Now so our overall results show that
1:00:33LGUs uh exhibit low to moderate AI
1:00:36readiness. So to leaning towards the the
1:00:39the left
1:00:42um
1:00:43the distribution peaks uh at a low to
1:00:46mid30s no you can see here in this uh um
1:00:50bell curve um meaning the average LGU
1:00:52has only partially built the foundations
1:00:55needed for AI adoption and there is a
1:00:58long right tail uh small but important
1:01:01group of higherforming LGUs of around 60
1:01:04the the score is already at 60 above 50
1:01:07na that are substantially more prepared.
1:01:11Okay. So let's try to dig deeper about
1:01:14the results. So looking at the six
1:01:16pillars um
1:01:19where do we actually perform well? It's
1:01:21actually in innovation. Innovation
1:01:23scores highest at 40.6 followed by
1:01:26economic foundations at 38.3 and data
1:01:29readiness at 36.1. So this
1:01:34so these reflect uh LGUs that have
1:01:37stronger income bases that have begun
1:01:39implementing judicial services. So in
1:01:42contrast uh governance scores only 27.8
1:01:45and skills making these the two most
1:01:48critical bottlenecks that currently that
1:01:51uh urgently need uh attention.
1:01:54Um next slide please.
1:01:57There are also substantial uh regional
1:02:00disparities. Um NCR leads by a very wide
1:02:04margin reflecting its concentration of
1:02:06digital infrastructure, skilled human
1:02:09capital and exposure to technologydriven
1:02:11sectors allow is actually benefiting the
1:02:14the region. At the other end we see
1:02:17Bakbarm
1:02:18uh Eastern Bisayas car cordelier
1:02:21autonomous region uh Cordelier
1:02:24administrative region rather and
1:02:25Mimaropa all average below 30 reflecting
1:02:29persistent development gaps in
1:02:30infrastructure skills and governance uh
1:02:33capacity
1:02:36looking at income class first class LGUs
1:02:39showed us highest scores um though and
1:02:42of course there's significant variation
1:02:44as you can see by the by the uh
1:02:47distribution of of the
1:02:50of the box curves. No, so meaning some
1:02:53are very ready while others are still
1:02:54lagging behind sometimes even uh uh
1:02:58lower than the second class uh
1:03:00municipalities. But what you can really
1:03:02see is as is income class falls from
1:03:04first to sixth uh AI readiness
1:03:07consistently declines. fifth and sixth
1:03:09class um LG use cluster towards the 30
1:03:14score in the index constrained by
1:03:16limited uh ICT investment, difficulty in
1:03:19attracting specialized staff and weaker
1:03:22institutional systems for data
1:03:24management and innovation.
1:03:27Okay, so let me skip this slide. This is
1:03:30just um going to look at more on the the
1:03:33very particular sector now. So on the
1:03:36digital infrastructure there is clear
1:03:39stratification by income level with
1:03:41higher income LGUs performing better
1:03:44overall.
1:03:46However um certain critical weaknesses
1:03:49particularly in internet sufficiency and
1:03:51ICT equipment density is actually
1:03:54persistent across all income classes. So
1:03:56while uh so it's common no it's not just
1:04:00a problem of the lower inclass groups
1:04:02but it's a common problem.
1:04:05On the next slide, we see some bright
1:04:07spots. Uh, for example, the availability
1:04:09of multiple network types is uh
1:04:12remarkably uniform across all in island
1:04:14groups, suggesting that basic
1:04:16connectivity coverage has become
1:04:18relatively standardized across uh the
1:04:21Philippines.
1:04:23Now, let's go to human capital and
1:04:25skills. Next, so human capital shows the
1:04:28sharpest divide of all six pillars. A
1:04:31substantial majority of first class LGUs
1:04:34have established ICT offices with
1:04:36dedicated personnel. But as income class
1:04:39falls, the capacity drops also
1:04:41dramatically. Many lower income LGUs
1:04:44have actually mentioned they have no
1:04:46dedicated ICT staff at all.
1:04:50uh across island groups, Bisayas shows a
1:04:52slight edge in ICT staff availability,
1:04:55but all island groups face the critical
1:04:57shortage in specialized ICT personnel
1:05:00and virtually none have formal digital
1:05:03skills development plans. So very few
1:05:05have digital skills development plans. U
1:05:09so some of them may be in it's embedded
1:05:11in their ICT development plan, but it's
1:05:14very critical to have a formal um
1:05:16digital ICT uh develop the skills
1:05:18development plan. Next slide please.
1:05:21Uh okay. So the innovation pillar is the
1:05:24strongest overall with LGUs uh across
1:05:27all incomes classes showing meaningful
1:05:30progress in digitalization of services
1:05:32which is good. However, adoption of
1:05:34emerging technologies such as uh our
1:05:37focus here AI remains critically low
1:05:40everywhere suggesting that current
1:05:42digitalization has not yet created the
1:05:44culture of technology exploration. So of
1:05:47course we are now exploring or looking
1:05:49at uh digital services but how to
1:05:53further improve that with the uh
1:05:55emerging technology of AI or that
1:05:57technology of AI that has already with
1:06:00us is actually not being uh explored.
1:06:04In the next slide, uh across island
1:06:06groups, the pattern is also strikingly
1:06:08similar. Uh strong digitalization
1:06:10progress, but in in terms of the
1:06:12adoption of emerging technologies, big
1:06:15data, AI, um the adoption is quite weak.
1:06:19So this uniformity suggests that the
1:06:21barrier is actually more nationally
1:06:24systemic uh rather than being um
1:06:26influenced by regional uh
1:06:28characteristics.
1:06:31The governance pillar reveals a striking
1:06:33paradox. No. So, LGU score well in
1:06:35digitalization plans, big data
1:06:37orientation on paper, but display
1:06:40critical weaknesses in uh actually
1:06:42allocating ICT budgets and developing
1:06:45medium-term planning. So, the intent is
1:06:48there. So some some actually have uh
1:06:50plans or they know of of the need to
1:06:53have a big data or or uh AI uh
1:06:57applications and and adoption of AI
1:07:00applications
1:07:02but the application and uh um
1:07:05implementation allocating budgets is is
1:07:08uh not there. So the intent is there uh
1:07:11across all income classes but putting
1:07:13your money where your plan is is
1:07:15actually not being uh done. the
1:07:17financial flow followth through is uh is
1:07:20not in the next slide. The same paradox
1:07:24um can be seen uh across island groups.
1:07:27Strong policy commitment to big data
1:07:29orientation contrast sharply with weak
1:07:32actual ICT budget allocation uh a
1:07:35pattern that is consistent uh
1:07:37nationwide.
1:07:39Data readiness is uh presents a mixed
1:07:42picture. Security measures and
1:07:44application usage are relatively strong
1:07:47but e-services availability the
1:07:49platforms actually needed to generate
1:07:52manage and share data AI applications is
1:07:55weak ac across some uh across all uh
1:07:58income income classes.
1:08:02Cyber security scores are exceptionally
1:08:05uniform nationwide with only 3.4 for
1:08:08points uh range separating uh uh all
1:08:11regions. Uh this indicates that cyber
1:08:14security has become a universally
1:08:16institutional priority um throughout the
1:08:19the Philippines uh uh local government
1:08:22uh system.
1:08:27So as expected um economic foundations
1:08:30consistently follow income class of
1:08:32course because that's your your classes
1:08:34are actually dependent on how much uh
1:08:37your your income is no.
1:08:40Yeah, higher uh income LGUs have
1:08:42substantially stronger fiscal autonomy,
1:08:44greater capacity for sub sustained
1:08:47technological invest technology
1:08:48investment while lower income classes
1:08:51face structural constraints in balancing
1:08:53local revenue against service
1:08:55expenditure needs.
1:08:57Interestingly uh in the next slide uh
1:09:00local revenue generation shows
1:09:02surprising uniformity across island
1:09:04groups. So most LGUs have developed some
1:09:06predictable uh local revenue base and
1:09:09this provides a starting point for
1:09:11digital investment even if it remains
1:09:14modest relative to what AI adoption uh
1:09:17requires.
1:09:19In the next slide, to complement our
1:09:21quantitative findings, we also conducted
1:09:24interviews with four uh local government
1:09:27uh representatives. Two from Mindanao,
1:09:30uh one from Luzon and um one from the
1:09:34Visayas, so spanning first class cities
1:09:36and uh municipalities. So what have we
1:09:39found? No. So in the next slide, uh all
1:09:42this is our findings. So all uh LGUs
1:09:45have no official short-term plans to
1:09:48adopt AI and the reasons are actually
1:09:50quite varied. So one cited the absence
1:09:52of skilled ICT workers. Another said
1:09:55their budget is uh limited uh
1:09:58prioritized for social services like
1:10:01scholarships and medical assistance. A
1:10:03third noted that AI providers have
1:10:05approached them about disaster risk
1:10:07reduction applications but no commitment
1:10:09has been made. So it's only up to the
1:10:11discussion level. Notably, one LGU
1:10:14shared um that st their staff the staff
1:10:17in the LGU are already informally using
1:10:20chat GPT and Grammarly for social media
1:10:24work. So, but that's at the individual
1:10:27level. No, so it's not institutionalized
1:10:29but this is some sign no a sign that
1:10:31grassroots AI use is already happening
1:10:34even without uh official plans.
1:10:37So in the next slide um what LGUs are
1:10:41actively pursuing is actually broader um
1:10:44digitalization. So online play payment
1:10:46platforms for local taxes driven by ease
1:10:49of doing business act and the programs
1:10:51like the OST's uh Starbucks for digital
1:10:54archives.
1:10:55AI is seen as a future step but not a
1:10:58current uh priority.
1:11:01Uh okay so we have uh five minutes
1:11:04remaining so that allows me some time to
1:11:07summarize our key findings. Um the
1:11:09Philippine uh LG use average 30 to 35 on
1:11:14the AI readiness index and that means
1:11:16it's a low to moderate uh preparedness
1:11:19income class is the most powerful
1:11:21predictor of readiness. uh NCR lead
1:11:24significantly with uh Barm, Eastern
1:11:27Visayas, CAR and Mimaropa um
1:11:32being at the lower end of of the scale
1:11:35and the skills pillar with an average
1:11:38score of 29 and the near total absence
1:11:41of ICT specialists across all income
1:11:43classes is actually the most critical
1:11:46gap to address.
1:11:49We have uh some policy recommendations.
1:11:52Uh
1:11:54uh first is about uh removing um
1:11:58regulatory barriers to infrastructure
1:12:00development
1:12:02and this means um eliminating the legis
1:12:05legislative franchise
1:12:07um
1:12:10requirement for connectivity providers,
1:12:12removing outdated laws like RA7925
1:12:16and uh the radio control uh law uh along
1:12:19with spectrum management reforms for 5G.
1:12:22G second
1:12:24uh is uh integrate uh AI and uh data
1:12:29literacy into education uh from K to2
1:12:32through higher education and uh Tesla uh
1:12:35building on RA119 uh27 and invest in
1:12:39teacher training uh capacity.
1:12:42Third uh is to establish a unified uh
1:12:46national AI strategy. But I know that
1:12:49the DOS has already released the
1:12:51national AI strategy. Um and this is uh
1:12:55actually a good step. But what is needed
1:12:57is actually designating um what would be
1:13:00the roles of the other agencies, the
1:13:02ICT, DTI, DO, DEP ed um and actually
1:13:08allowing for flexibility within each
1:13:10sector to develop tailored policies to
1:13:13help support the the LGUs in the
1:13:15implementation of uh in the adoption of
1:13:18AI.
1:13:19Fourth, mandate a minimum ICT budget
1:13:22allocation and if um this is uh just as
1:13:25as much as possible no so for 2% for LGU
1:13:29budgets um for for digitalization and
1:13:33then rising increasing it uh eventually
1:13:35to 3%. And the link the compliance to
1:13:38eligibility for national grants and
1:13:41performance-based uh transfers.
1:13:44Fifth is establish a national data
1:13:46governance framework um under the PSA to
1:13:50with standards for data generation,
1:13:52collection, sharing and privacy plus a
1:13:54centralized data portal for authorized
1:13:57researchers and regulators.
1:14:00Sixth is to create a national LGU AI
1:14:02readiness fund uh with equity based uh
1:14:05allocation. So 100% grants for fourth
1:14:08through sixth class LGUs with sliding
1:14:11scales for higher income classes.
1:14:14conditional on the credible
1:14:15implementation plans and commitment to
1:14:18knowledge sharing. So uh this is a means
1:14:20of ensuring that the um lowerass LGUs
1:14:25would be able to uh catch up with the
1:14:27higher class uh LGUs and um with this uh
1:14:31we end uh our presentation. Thank you
1:14:34very much uh for your attention and we
1:14:37welcome your comments and suggestions.
1:14:39Thank you.
1:14:41>> Thank you so much Dr. Francis for your
1:14:43insightful presentation. So presentation
1:14:49potential AI governance
1:14:53readiness across our LGUs
1:14:57clear disparities across regions and
1:14:59income levels
1:15:02evenly distributed readiness
1:15:05AI. All right. So once again thank you
1:15:08po Dr. Francis for your presentation.
1:15:10Soon we have 1,5 participants on Zoom.
1:15:15Uh we have reached our maximum capacity.
1:15:17So thank you so much po for joining. We
1:15:20also have 417 viewers on Facebook live.
1:15:24Thank you all for joining us today. So
1:15:26for those who still want to catch up no
1:15:28you can continue watching uh uh the live
1:15:31stream on the pets and Facebook pages.
1:15:35We also invite you to share your
1:15:37thoughts uh using uh our official
1:15:40hashtags # petsvents
1:15:44#ip
1:15:45and # futureof work. For our uh Facebook
1:15:50live viewers, you may also let us know
1:15:53where you're watching from. No, just
1:15:54drop your location in the comment
1:15:56section. So, let's give a quick shout
1:15:59out to some of our uh FP viewers joining
1:16:02us today. First we have uh Silver Gaton.
1:16:06I hope I'm pronouncing your surname
1:16:08right now. From the Philippine
1:16:10Statistics Authority. Thank you for
1:16:11joining. We have Sire De Laente from the
1:16:15University of Antique.
1:16:18Good morning. Uh we have Lao Catod from
1:16:21Land Bank head office. Thank you for
1:16:24joining. Uh Vandom Boromeo Gallardo from
1:16:27First Lagona Electric Cooperative
1:16:30Incorporated, Lumband. May viewers ta
1:16:34from the National Bureau of
1:16:35Investigation,
1:16:37House of Representatives,
1:16:39Northern Northeastern Mindanao State
1:16:42University and uh Accidental Menuro
1:16:45State College and Gingo City United
1:16:48Colleges
1:16:50for joining us today and good morning.
1:16:54Please stay tuned until the end of the
1:16:56program FB live viewers. uh you can
1:16:59receive your e certificate through our
1:17:02replay on demand.
1:17:04Meanwhile, for our Zoom participants,
1:17:08e certificates after completing the
1:17:10feedback survey. So
1:17:13team links for this after the discussion
1:17:16and of course
1:17:19chance
1:17:20exclusive pads merchandise. So kamatong
1:17:24pads merch items today. So make sure to
1:17:28stay until the end of the program para
1:17:30may chance po manalo. All right. So we
1:17:33now move to our third and final study
1:17:37complementing the governance
1:17:39perspective. This study is titled impact
1:17:42of artificial intelligence on the labor
1:17:44market. Authored by senior labor and
1:17:46employment officers Chelsea Nicole
1:17:49Pineda and yours truly Marvin
1:17:51Kabangunai. The study examines how AI
1:17:54adoption is reshaping jobs, skills, and
1:17:57work experiences across key sectors.
1:18:00Findings show that while AI enhances
1:18:02productivity and creates demand for new
1:18:05occupations, it also introduces risks
1:18:08such as job displacement and skills
1:18:10mismatches, highlighting the need for
1:18:13coordinated policies on workforce
1:18:15development, reskilling, and the ethical
1:18:17use of AI. This will be presented by
1:18:20Miss Chelsea Nicole Pineda. Miss Pineda
1:18:24holds a master's degree from the
1:18:25University of the Philippines DMAN and a
1:18:28bachelor's degree from Deasal College of
1:18:30St. Pine. Her work focuses on the future
1:18:33of work, labor markets, migration and
1:18:37employment policy and she also teaches
1:18:39courses on ashan and migration studies.
1:18:42Let us now welcome Miss Pineda.
1:18:48Thank you for the introduction, Morbin.
1:18:50I will just share my screen or my slide.
1:18:55So, can I confirm if my screen can now
1:18:58is now visible?
1:19:04Okay. So, thank you very much for the
1:19:06introduction, Marvin. And before I start
1:19:08my presentation, I'd like to invite
1:19:10everyone to open their Messenger apps.
1:19:12And if you can see satas po search meta
1:19:16AI or search. So with this we can see
1:19:19that AI can now easily be accessible by
1:19:23everyone and anyone. And you don't have
1:19:26to have at least specialized AI skills
1:19:28to know or use artificial intelligence.
1:19:31But let's contextualize our discussion
1:19:34on how AI is being implemented in the
1:19:37Philippine labor market. Our study
1:19:39impact of artificial intelligence on the
1:19:42Philippine labor market was conducted
1:19:44back in 2024 where policy discussions
1:19:47are being paved on how AI is being used
1:19:50in certain industries and what is it
1:19:52what is its impact in the labor market.
1:19:55So we surveyed around 33 uh sorry 54
1:20:00employers and around 300 employees in
1:20:03three different sectors to see how AI is
1:20:07being implemented and what are the
1:20:09current realities of the labor market
1:20:12back in 2024.
1:20:14So right now we are currently in the
1:20:16fourth industrial revolution where there
1:20:18are transformative shifts in the world
1:20:20of work altering how tasks are being
1:20:23performed and how job roles are being
1:20:25defined for the 21st century. There are
1:20:28two sides of the AI debate in the world
1:20:31of work. First on the one hand AI
1:20:34increases productivity efficiency and at
1:20:37least it reduces repetitive tasks.
1:20:40However, on the other hand, it also has
1:20:42negative effects. A lot of people are
1:20:45concerned on the job displacement that
1:20:48could occur while certain industries are
1:20:51also saying that AI could impact the
1:20:54cognitive ability of the workers because
1:20:56it is now um taking up the repetitive
1:21:00task. So, more workers are now using
1:21:02their cognitive functions to perform
1:21:04their tasks. In 2024, we surveyed
1:21:08employers from three different
1:21:10industries, namely the IPBPM,
1:21:12manufacturing, and the banking and
1:21:14finance industries. Because back in
1:21:172024, we've seen that they are at the
1:21:20forefront of AI adoption here in the
1:21:23Philippines. And among our surveyed
1:21:25respondents, majority are already
1:21:27implementing AI within their
1:21:30enterprises. and around 49% are not
1:21:35implementing AI but they are currently
1:21:38looking into how AI can be implemented
1:21:41within their enterprises. So why are
1:21:44employers adopting artificial
1:21:46intelligence? First is because of its
1:21:50cost uh productivity, cost efficiency
1:21:53and quality assurance. Different
1:21:54industries have different priorities due
1:21:57to their specific grade tasks and
1:22:00because certain industries like the ITPM
1:22:04have customer service and clients
1:22:07overseas, it is important for them to
1:22:09keep on the tide and be updated with the
1:22:12current trends in the market and it is
1:22:15important for them that they uh optimize
1:22:18their operations to their overseas
1:22:21clients. Meanwhile, in the manufacturing
1:22:23sector, since majority of our
1:22:26manufacturing companies here are
1:22:27subsidiaries of international companies,
1:22:30they adopt AI because it is mandated or
1:22:33introduced by their inter international
1:22:35headquarters. Meanwhile, banking and
1:22:38finance institutions here in the
1:22:40Philippines are using AI within their
1:22:43client services because they want to
1:22:45stay competitive in the market and they
1:22:48want to improve their customer
1:22:50satisfaction.
1:22:53How is AI being implemented in these
1:22:56industries? So for uh different
1:22:58industries, they have different use
1:23:00cases. For um the banking and finance,
1:23:04they are using virtual agents so that
1:23:06their clients can tune can can
1:23:08communicate to these virtual agents or
1:23:10chatbots to answer their inquiries.
1:23:13Meanwhile, um within their internal
1:23:15operations, there are implementing AI to
1:23:18increase their fraud management
1:23:20capabilities and also um to provide a
1:23:25deeper and sharper analysis of their
1:23:27clients credit scores. And in the
1:23:31context of the manufacturing sector,
1:23:33they are using AI to uh increase their
1:23:36predictive repair and virtually inspect
1:23:39their machinery. Since uh the machines
1:23:43of these manufacturing companies are
1:23:46very big and are very complex and I
1:23:49think one of at least a common use case
1:23:52among our enterprises is the applicant
1:23:55tracking system. Traditionally HR
1:23:58professionals would um match the skills
1:24:01and competencies of the job seekers
1:24:04against the job requirements of their
1:24:06posted vacancies. But because of
1:24:08applicant tracking systems of online job
1:24:12sites, they are now eliminating these
1:24:15manual task and they now have more time
1:24:18to at least assess the candidates
1:24:21competencies and skills.
1:24:25Despite having no AI regulation at the
1:24:28national level, companies who have
1:24:31adopted AI in their business operations
1:24:34are exercising self-regulation in spirit
1:24:37of responsible business conduct and
1:24:39corporate governance to protect their
1:24:42core business interests and also uphold
1:24:44their trust with their clientele. So
1:24:46this denotes the impact of digitization
1:24:49and business transformation in the 21st
1:24:51century. Some companies have opted to
1:24:54issue a separate guideline for the use
1:24:57of AI in their uh workplace that
1:25:00incorporates values such as
1:25:02transparency, ethical and responsible
1:25:05use. And then some companies are using
1:25:09their existing IT guidelines because
1:25:11this also coincides with their data
1:25:13privacy regulations and compliance and
1:25:17their cyber security infrastructure.
1:25:19Certain companies who hold confidential
1:25:23and sensitive information uh incorporate
1:25:26a provision that the use of AI should
1:25:30only be for company related task or
1:25:33within the bounds of the agreement with
1:25:35their clients. especially that in the
1:25:37ITBM sector, certain accounts are
1:25:40requiring their uh offshore operations
1:25:44here in the Philippines to use a certain
1:25:46AI agent or an AI system.
1:25:51So that was the perspective of the
1:25:53employers and also we will now tackle
1:25:56the affected workers due to AI adoption
1:25:59adoption. So among the surveyed
1:26:02employers, we gathered that only 6% of
1:26:06the surveyed employers experienced
1:26:08employee reduction due to AI adoption.
1:26:11The highest number of affected workers
1:26:13was from the ITDPM sector where 31 to 40
1:26:17employees were affected because of AI
1:26:20adoption. And just to contextualize,
1:26:23some of these displaced workers are from
1:26:26the information technology and
1:26:28information management department.
1:26:32Before these employees were uh provided
1:26:37with se uh with separation
1:26:40packages, some companies opted to first
1:26:43transfer their employees to a different
1:26:46unit or different department within the
1:26:48company because their existing skills
1:26:51also matches certain uh demands within
1:26:54the business units of this enterprises.
1:26:57While those who were um affected by the
1:27:01AI adoption were referred to other
1:27:03companies, others were referred to the
1:27:05dole adjustments measures program.
1:27:08Others were processed for the SSS
1:27:10unemployment insurance and uh certain
1:27:14employees were also provided with
1:27:17separation packages.
1:27:21Majority of our surveyed respondents
1:27:24also told us through a survey form that
1:27:28their workers are organized. Hence some
1:27:32of these organized uh employees raised a
1:27:36discussion to the management regarding
1:27:39the adoption of AI within their
1:27:43enterprises. And some of the concerns
1:27:45raised by these employees are skills and
1:27:48training programs, the job security of
1:27:50the workforce, the productivity and
1:27:53mental health. And I'd like to highlight
1:27:55that the mental health is correlated
1:27:56with the job security because we still
1:27:59back then they are still not aware of
1:28:01the real impact of AI and that the
1:28:04prevailing notion that AI adoption would
1:28:07result into job loss and that's why
1:28:09they're have poor mental health when
1:28:12they learn that AI will be adopted and
1:28:14raise concerns regarding their security
1:28:17of tenure. Some um you uh some organized
1:28:20employees also raised regarding
1:28:22renumeration concerns if they would have
1:28:25an increase or a decrease. Next, in the
1:28:28manufacturing sector, it was very
1:28:30specific that they raised their physical
1:28:32safety of using AI in the uh production
1:28:37line because they are still um not
1:28:39clearly adept to how AI will be used as
1:28:42a complement or as a tool to implement
1:28:45predictive defense and virtual uh
1:28:47inspections. And while others are non uh
1:28:50other establishments have no formal or
1:28:53organized employees, these employees
1:28:57have also raised direct concerns to
1:29:00their supervisors regarding the adoption
1:29:02of AI or at least in town hall
1:29:05assemblies where there is a periodic or
1:29:09quarterly dialogue between the
1:29:11management and the employees.
1:29:15Uh on this slide I will present the
1:29:18concerns of the employees among the
1:29:21surveyed uh employees because we want to
1:29:24triangulate the claims of the employers
1:29:27and as well as the experience of the
1:29:30employees who are already using AI in
1:29:33their work. So based from our survey
1:29:36among the employees they are primarily
1:29:39concerned on the workload because of AI
1:29:42adoption. So this is contrary to what is
1:29:46uh being said that AI eliminates
1:29:48repetitive task. To contextualize again
1:29:52because of AI adoption they feel that
1:29:54their workload has increased because
1:29:56they have to check what the AI has
1:30:00provided them. So for them it increases
1:30:02their time and their taskings to double
1:30:06check how if the AI has provided correct
1:30:09and accurate inputs. Meanwhile, it is
1:30:13still a concern among our employees
1:30:15regarding their job security. So, it is
1:30:18still in the back of their head that in
1:30:20the future I might still be displaced
1:30:23because of AI adoption. And others are
1:30:26also concerned of use of AI in the
1:30:29decision making. Because when HR
1:30:31managers are using AI to have a decision
1:30:36on the security of tenure of the
1:30:39employee, it takes away the qualitative
1:30:42um assessment on the employ on the
1:30:45employees and so that they are concerned
1:30:48of how AI is being used especially in
1:30:50critical decision making that involves
1:30:53employees security of failure
1:30:57and at their individual level. These
1:31:01employees who are working with AI has
1:31:06also raised concerns regarding the
1:31:09privacy and security of the data that
1:31:11they are working on. Most especially
1:31:14that two of the sectors that we have
1:31:16surveyed are handling sensitive and
1:31:18personal informations of local and
1:31:21overseas clients. Meanwhile, it is also
1:31:24an interesting finding of this study
1:31:26that there are certain employees that
1:31:29has raised concerns on job satisfaction.
1:31:34And while AI has been found to affect
1:31:38the security of tenure of uh employees,
1:31:41it also generated demand for AI skills
1:31:45such as AI and machine learning
1:31:46specialists, AI engineers, data
1:31:49analysts, AI ethicists and prompt
1:31:52engineers. So let me highlight the
1:31:54demand for AI ethicis because it
1:31:56reflects the discussion that AI should
1:32:00be responsibly and ethically used and
1:32:03certain companies are already devising
1:32:05strategies on how to formulate and
1:32:08design their AI systems that coincides
1:32:11with this very important value of
1:32:14ethical and responsible use. especially
1:32:16certain um AI systems can be used for
1:32:21illicit activities or in non-company
1:32:25related tasks that involve the personal
1:32:27and confidential information of clients.
1:32:32We also tried to look on the impact of
1:32:35AI on the wages across these sectors and
1:32:38majority of our respondents said that AI
1:32:41adoption did not increase their wages
1:32:44but notably 12% told us that AI adoption
1:32:48indeed increase their wages to which the
1:32:51banking and finance has the highest
1:32:53self-reported num uh increase in wages
1:32:57ranging from 45,000 to at least 55,000
1:33:01And the lowest increase that was
1:33:04reported was in the IPDPM sector which
1:33:07was just below 15,000. And notably there
1:33:12are no
1:33:14reported increases from our
1:33:15manufacturing employees which highlights
1:33:18that the impact of AI is unequal among
1:33:21employees and in different sectors.
1:33:26So let me just also provide the current
1:33:28snapshot of what is uh the pay skills of
1:33:34AI occupations here in the Philippines.
1:33:37So from a a job site we can see that the
1:33:41a data analyst can make at least 32,147
1:33:46per month as of April 2026 while data
1:33:49engineers can earn at least 66,000
1:33:53per month.
1:33:55On the other hand,
1:33:58AI and ML machine learning engineers
1:34:01could earn to at least 66,952
1:34:05per month this April 2026 while AI
1:34:08training specialists could also earn
1:34:1147478
1:34:13per month. So this highlights the demand
1:34:15for AI occupations as well as the
1:34:19availability of skills here in the labor
1:34:21market. So this could also be a jump off
1:34:24of our educational institutions on how
1:34:27they can design their training and
1:34:29curriculum pattern towards the demands
1:34:32of the current industry.
1:34:36Now because of learning AI skills,
1:34:38majority of our respondents confirmed
1:34:41that because AI has provided them with
1:34:46technical skills that enable them to use
1:34:48AI or make and build AI from the
1:34:52scratch. They have now aspirations for
1:34:54career mobility. And we can see that
1:34:57there is a high demand for data science
1:35:00and uh analytics. And when equipped they
1:35:03will be certain that they will move from
1:35:06a different company that offers um
1:35:09better compensation packages.
1:35:13Regarding task performance, AI has
1:35:16increased at the pace of the task or uh
1:35:21when they are performing them.
1:35:23Meanwhile, 25% of our surveyed employees
1:35:27told us that because AI has took the
1:35:30repetitive and manual task, they can now
1:35:32more focus on cognitive taskings. Well,
1:35:37there is also a self-reported
1:35:39um 9% that AI did not increase the pace
1:35:43of performing task and uh small
1:35:47percentage only 7% reported that AI
1:35:51increased the chances of their workload
1:35:55and notably this um rise is within the
1:35:59IPVPM sector
1:36:04in terms of job satisfaction and job
1:36:06performance. Uh across our
1:36:10across our surveyed employees, around
1:36:1349% agreed that AI improved their job
1:36:16satisfaction and 46%
1:36:20told us that AI has improved their job
1:36:23performance. So in terms of visibility
1:36:27across different industries, majority of
1:36:30our respondents are from the ITBM. Hence
1:36:33they are satisfied and with their
1:36:36current jobs because AI was introduced
1:36:39while they are also um looking into an
1:36:43increase in their job performance
1:36:45because AI has helped them to do some of
1:36:48the manual and repetitive task and help
1:36:51them improve as employees.
1:36:55Now, regarding the employee sentiments
1:36:57of using work, at least 71% of our
1:37:01surveyed respondents was glad that AI
1:37:05was introduced in their workplace, which
1:37:08is generally across all of the sectors
1:37:11that we uh surveyed. Meanwhile, 69% do
1:37:16not feel that they are concerned that AI
1:37:18will replace them because they finally
1:37:21know how AI can be used as a tool within
1:37:24their job occupations and how it and how
1:37:27it could be used to complement the work
1:37:30that they are doing. Meanwhile, it is
1:37:32still notable that there are employees
1:37:35who are still concerns that AI will
1:37:38replace them in the future, especially
1:37:41that they know um they need to upskill
1:37:47and increase their portfolios to stay
1:37:50competitive and employable in the labor
1:37:53market. Now that more enterprises or
1:37:56more enterprises within their industry
1:37:59are adopting AI.
1:38:03So while AI adoption has increased in
1:38:05different sectors, it also raises
1:38:07initial concerns on its adverse impacts
1:38:10on the education and employment
1:38:12landscape. So there is still concerns on
1:38:15displacement in the workplace and reli
1:38:18over reliance on AI could lead to a
1:38:21decline in soft skills especially on
1:38:24problem solving, critical thinking and
1:38:27persontoperson communication. AI could
1:38:30also be used for illicit activities as
1:38:32it could generate images, audio and
1:38:36videos that could cause spread of
1:38:39misinformation, misrepresentation
1:38:42and illegal activities because at
1:38:44present there is no law that
1:38:47criminalizes the use of AI only. There
1:38:50are discussions at the legislative um on
1:38:53how they could criminalize the use of
1:38:55deep fake videos or audios.
1:39:01Next, over reliance could still also
1:39:04lead to poor decision making and loss of
1:39:07human intervention in AI generated
1:39:10outputs. And it is also important to
1:39:13highlight that data quality and the
1:39:15reinforcement of biases could also be um
1:39:20used in designing AI models because it
1:39:23could further exacerbate discrimination
1:39:26and it and inequality. Meanwhile, since
1:39:30the nature of AI is digital
1:39:32technologies, data owners may not be
1:39:34aware that their personal and sensitive
1:39:37information is being used to build and
1:39:40train AI models. And this poses a danger
1:39:43especially that um you did not provide
1:39:46your consent on your data being used as
1:39:50a model for building AI technologies or
1:39:53systems.
1:39:56So overall our study recommends first
1:39:59government coordination. So there should
1:40:01be a convergence with national
1:40:03government agencies especially for
1:40:05employment facilitation on job
1:40:07opportunities related to AI. In terms of
1:40:10skilling programs, there is a need to
1:40:12fasttrack the development of training
1:40:14regulations and on data science and
1:40:16analytics and competency standards for
1:40:19other AI related fields and resource
1:40:21sharing between public and private
1:40:23institutions on the education and
1:40:25training curriculum. But currently Tesla
1:40:28has already promulgated training
1:40:29regulation and compet competency
1:40:32standards for data science and
1:40:35analytics.
1:40:37In terms of partnership, there should be
1:40:39a tripertite helix where industry,
1:40:41government and academ should converge to
1:40:44promote the the dissemination of
1:40:47skilling programs
1:40:49on AI and other AI related skills. And
1:40:53at the national level, there should be
1:40:55support on the creation of a national
1:40:57level oversight NP on AI to enable an
1:41:01inter agency comprehensive plan to
1:41:03address labor and employment concerns.
1:41:06Especially that AI cuts across different
1:41:09policy areas of labor and employment,
1:41:12economics, competitive policy and as
1:41:15well as infrastructure
1:41:17and information technology.
1:41:23We should also pursue benchmarking
1:41:25activities with local industries that
1:41:28already adopted AI so that we can see
1:41:32how is it being implemented and what
1:41:35best practices could be adopted from
1:41:37other industries. Meanwhile, our
1:41:40foreigner counterparts or other
1:41:41countries can also be used as benchmark
1:41:45practices to see at least how they
1:41:47develop their human resources to prepare
1:41:50the future entrance in the labor force
1:41:53for an AI powered labor market. In terms
1:41:56of soft skills, we recommend to
1:41:58reinforce the implementation of 21st
1:42:01century skills to ensure that the labor
1:42:04supply is aligned and that we could also
1:42:09see an increase of um these future
1:42:12entrance in the labor market there that
1:42:15they are still competitive and
1:42:16employable not only here in the local
1:42:19labor market but as well as the
1:42:21international so that they could be
1:42:24adaptable to to the technological
1:42:27innovations uh at the global stage most
1:42:31especially that technology moves at a
1:42:34fast pace. Next, we also recommend for
1:42:38industry specific job fairs to provide
1:42:41employment opportunities to those job
1:42:45seekers that have trained with AI and
1:42:47other AI related skills. And lastly, we
1:42:51recommend to increase the steam of
1:42:53science, technology, engineering, and
1:42:55mathematics graduates in labor supply
1:42:57institutions in anticipation of the
1:43:00future labor demand for AI and AI
1:43:03related occupations. So, this would help
1:43:06the our uh this would help our
1:43:09enterprises
1:43:11innovate and stay competitive in a
1:43:14global race for AI adoption.
1:43:18So let me end my presentation with a
1:43:21remark that AI must be designed to
1:43:23respect the fundamental fundamental
1:43:25human rights and should complement human
1:43:28work and not replace individual workers
1:43:31within enterprises.
1:43:34If you want to see the full results of
1:43:35the study that covers the immediate and
1:43:38perceived impact of artificial
1:43:40intelligence in the labor market, please
1:43:42scan the QR code as shown. Thank you for
1:43:46listening and I look forward to the
1:43:48question and answer portion.
1:43:52>> Thank you very much uh Chelsea for your
1:43:54presentation. So as discussed speakers
1:44:02tangible benefits workplace in improving
1:44:06uh efficiency and job performance
1:44:09at the same time valid concerns then
1:44:12from workers especially
1:44:14job security skills requirements and
1:44:17ethical use of AI. Now that we have
1:44:20heard the presentations and insights
1:44:22from our speakers, we'll now move to our
1:44:25open forum. So for our Zoom
1:44:28participants, you may send in your
1:44:29questions through the Q&A box. For our
1:44:32Facebook live viewers, poin
1:44:36questions uh s comment section. So at
1:44:39this point, may we request our panelists
1:44:42to kindly turn on their cameras as we
1:44:45proceed with the Q&A session. So, uh,
1:44:48Professor Alexis, Dr. uh, Francis and,
1:44:52uh, Chelsea. Okay. So, we're very, uh,
1:44:55encouraged by the level of engagement in
1:44:57today's discussion. We've received more
1:44:59than 30 questions from Zoom along with,
1:45:02uh, several comments from Facebook. And
1:45:04as of the moment, we have about 1,5
1:45:07participants on Zoom and more than 468
1:45:11viewers on our Facebook live. So given
1:45:13our limited time uh we may not be able
1:45:16to address all questions during the
1:45:18session. However uh for those uh who are
1:45:22unable to take up live we will uh
1:45:24request our presenters to respond via
1:45:27email. So uh now uh let's proceed with
1:45:31uh the questions. No uh let me read the
1:45:33first question. Uh this is from
1:45:36professor Nancy Eeria of uh University
1:45:40of Stomas. Um this is directed to
1:45:44professor Alexis. So uh you mentioned
1:45:47that artificial intelligence or AI may
1:45:50lead to a shrinking middle class. Could
1:45:52you briefly explain how this might
1:45:55happen and what policies or action can
1:45:58help prevent or minimize this? Professor
1:46:02Alexis.
1:46:03>> Yes, thank you Marvin and and thank you
1:46:05to um to Professor Nancy for posting
1:46:08this question. In fact, I wrote a few
1:46:10down. So for the sake of time if you
1:46:12want me I can give quick comments but
1:46:15let me talk about that one. So we do
1:46:17have um inequality rising and we see
1:46:20that if you look at the income
1:46:22distribution the middle class starts to
1:46:25disappear. AI will only accelerate that.
1:46:28So what is happening it's a very naive
1:46:31example but think of those the
1:46:34aristocrats in the old times they
1:46:36control the land and everybody worked
1:46:38for those land owners now we have the
1:46:41people that control the AI machines that
1:46:45create power that everybody needs to use
1:46:47clo now needs to use so it's those that
1:46:50control those machines and of course
1:46:52control means those that already
1:46:54invested in them like sovereign wealth
1:46:55funds and big funds all the revenue will
1:46:58be going into those AI machines.
1:47:02The the machines will be doing the work
1:47:03that human beings used to do and the
1:47:06rest of us will just exist.
1:47:08It's a very extreme but I think that
1:47:10captures why I think the middle class
1:47:13will start to disappear. That said, I
1:47:16will comment about another question.
1:47:18Somebody asked about nurses. That's very
1:47:20interesting because as we we we end up
1:47:25living a longer life which is great
1:47:28there is more need for support for
1:47:31nurses and that cannot be replaced by
1:47:33AI. So while there are some doors
1:47:35closing there are some others that are
1:47:37going to open and I think the an area
1:47:40where the Philippines will benefit a lot
1:47:42will be by exporting nurses. It's
1:47:46unfortunate because you have to export.
1:47:48They have to leave their families to
1:47:49earn a living. But that's an area where
1:47:51like you can have a you can still have a
1:47:54middle class. Somebody talked about
1:47:56large language models. Maybe I can
1:47:57discuss this later on. But they need to
1:48:00upgrade the infrastructure on on water
1:48:03on energy. So we talk about
1:48:05infrastructure now. We need workers. So
1:48:08there will be some opportunities. It's
1:48:10not like everything will be wiped out
1:48:12but there will be much fewer. Thank you.
1:48:17Thank you so much uh professor Alexis
1:48:19for your response and thank you also to
1:48:21professor Nancy for your question. So um
1:48:25for our next question this is from Miss
1:48:27Jennifer Garcia of the University of
1:48:30Matti. Um the question is how will AI
1:48:33impact academic institutions? What
1:48:36abilities will be required um of
1:48:39professors and educators? um from Miss
1:48:42Maria Lordis Mendulsa from the House of
1:48:44Representatives um on Facebook has a
1:48:47related question on this stating that
1:48:49the prevalent use of AI will exacerbate
1:48:52the marginalization of the poor. What
1:48:55interventions in our academic system are
1:48:57necessary to uh help the poor students
1:49:00cope with this trend. So to summarize
1:49:03the question um the question is uh how
1:49:06is AI expected to reshape academic
1:49:10institutions uh and the role of our
1:49:13educators and what should schools uh do
1:49:16to ensure that poor students are not
1:49:18further marginalized. So this uh can be
1:49:22answered by any of our presenters.
1:49:24Perhaps we can start again with uh
1:49:26professor Alexis uh to be followed by
1:49:29Sir Francis and then Chelsea.
1:49:32Thank you. AI has changed everything in
1:49:35education. Our stu and it's not just AI.
1:49:39I I see that postcoid everything changed
1:49:43everything. So our kids learned to be
1:49:45online with COVID and they see this is
1:49:47fine. Now it's normal to take classes
1:49:50online to interact with their friends
1:49:51online to find activities games to do
1:49:53online. Second thing was a discovery of
1:49:56reals
1:49:57that we didn't have before covid. So now
1:50:01uh they their attention and sometimes we
1:50:04find ourselves going from one real to
1:50:06the other. If we don't get the
1:50:07information in 5 to 10 seconds we cannot
1:50:10focus. So kids are not reading books.
1:50:13You cannot convince them to open a
1:50:15textbook to spend time to understand
1:50:17something before they do it. And and
1:50:19then AI came and changed everything
1:50:21because AI provides the answer. So to
1:50:23answer the the first challenge is how do
1:50:25we teach the kids to think and leverage
1:50:28AI to build things
1:50:31because if if you use AI just to give
1:50:33you the answer to a problem and you just
1:50:37paste it, you will be the first one to
1:50:39be obsolete in the labor market. But if
1:50:42you find a way to think and then you use
1:50:45AI to build, then you will be ahead of
1:50:48the curve.
1:50:50And this is very important. And I want
1:50:52to I don't want to take time from my
1:50:54fellow panelists but this is very
1:50:55important. You can be a very creative
1:50:58person but you didn't know coding. That
1:51:01was a very big barrier. So you never
1:51:04tried that. You were a great designer
1:51:06but you never figured out because you
1:51:08didn't know how to use software to
1:51:10design things. Now with AI we don't need
1:51:13to know coding. We don't need to know
1:51:14how to use AutoCAD and 3D rendering. If
1:51:17we we can discover we have skills we
1:51:20didn't know we had before. So that
1:51:22creates opportunities. So the challenge
1:51:23for education is how do we we try to
1:51:26create those opportunities and teach
1:51:28those kids how to discover it uh
1:51:30themselves and this is you know this is
1:51:33changing things.
1:51:36So uh so it's a challenge for academics
1:51:39to find ways to incorporate AI into the
1:51:42classroom not just to penalize students
1:51:44from using AI to cheat but to build
1:51:47things. It's a challenge at a different
1:51:49level. um on you know we do research and
1:51:53our quality is based on where we publish
1:51:56that has changed so I think that's
1:51:58destroyed I can write 15 papers in in
1:52:01one day I can submit to journals I can
1:52:05pass them through so many IA tools
1:52:08referees will be overwhelmed by the
1:52:10quality and the quantity so they will
1:52:13not be able to do the job right my my
1:52:17merit review which is basically how I
1:52:19get increase in my salary is based on my
1:52:21publications but that's so it messes up
1:52:23everything so I have no hope for that I
1:52:26hear people going to study economics PhD
1:52:29and I tell them you know why I mean it's
1:52:32just changing everything very fast but
1:52:34what we need to focus is how do we give
1:52:35these tools to the kids and uh I I'll
1:52:40give you one example because that goes
1:52:42to another question what type of skills
1:52:44do we need to teach them I was teaching
1:52:47an executive course for the kadar
1:52:49investment authority for the new hires.
1:52:51These are people very bright.
1:52:54Katar Investment Authority is a so
1:52:56wealth fund. Very bright. They studied
1:52:58finance, computer science, economics,
1:53:01history, very different backgrounds.
1:53:04And I asked them to build an algorithmic
1:53:07trading model that uses network analysis
1:53:09to create trading strategies. They were
1:53:12looking at me like I was crazy because
1:53:14this thing didn't make sense to use
1:53:16neural networks. And I explained to them
1:53:18what it is. And I said just use AI and
1:53:21explain step by step. Now two girls
1:53:26finished the project and then they were
1:53:28able to build. Why? Because they treated
1:53:32AI as a 5-year-old kid. They were able
1:53:35to communicate to explain in a very
1:53:39articulate way hold it by the hand break
1:53:42the task into steps build one and then
1:53:46go to the next one. And then AI
1:53:48understood what they were trying to do
1:53:50and then they could do so many different
1:53:52things. Some other people that studied
1:53:54finance and they had masters they told
1:53:56AI go and do this for me but they they
1:53:59lack the communication skills
1:54:02and they failed. So soft skills are
1:54:04actually becoming even more important
1:54:06now that the technical barrier has
1:54:08broken. So to answer the challenge for
1:54:10education is twofold. We need to show
1:54:13them how what is AI and how they can use
1:54:15that. And it's hard for us because we
1:54:17are not ahead of the curve. We are so
1:54:19behind. But also to emphasize the
1:54:21importance of soft skills at the same
1:54:24time. Thank you.
1:54:27>> Thank you so much Prof. Alexis. Uh
1:54:30perhaps uh Dr. Francis has an additional
1:54:33insight.
1:54:34>> Uh thank you Marvin. No, so I agree with
1:54:37what Professor Alexis has said. No. soft
1:54:39skills are actually very important.
1:54:42Teaching uh grit uh to to students and
1:54:45enhancing their uh creativity and um uh
1:54:51uh uh yeah the their their creativity
1:54:54and critical thinking. Um but I think
1:54:57it's also important to understand um to
1:55:00see uh from the point of view of the
1:55:02schools no so of the administrators
1:55:05um because schools are as prof uh
1:55:08professor Alexis has mentioned no so co
1:55:12has uh really um thrown a wrench to how
1:55:17schools are actually being viewed but um
1:55:21at least from my experience when I was
1:55:23in schools has actually taught as um
1:55:26social skills. Now the so these are
1:55:28skills that you don't really um learn by
1:55:32through through the digital um and has
1:55:35actually gone down through the digital
1:55:36age. But um so this the schools are
1:55:39actually a a brewing pot of of um
1:55:43enhancing your emotional intelligence
1:55:45and and enhancing your your the the soft
1:55:48skills. And so I think this is where
1:55:52schools can actually pivot. No. So
1:55:54looking at a way for schools to be able
1:55:56to um enhance the emotional
1:55:59intelligence, discipline, grit of
1:56:02students beyond academics. No, because
1:56:04of of course uh AI will be there and um
1:56:08what we learn in lessons will be there.
1:56:10But I think this is something that what
1:56:12uh something that schools can do. Second
1:56:14is also um schools are actually a way
1:56:19for marginalized students to actually
1:56:22have access to um uh infrastructure. So
1:56:28different infrastructure um I remember
1:56:30when I was in school I we didn't have a
1:56:32lot of resources but the library has a
1:56:35lot of resources. No, so there are a lot
1:56:36of books. Um internet is uh fairly
1:56:39available. No, of course you have to
1:56:41line up for all those big computers that
1:56:44are available only for 1 hour, but it it
1:56:47gives you a chance to write your paper
1:56:49and so but again so the schools are
1:56:52actually um the means for the for the
1:56:56students who don't have access to AI to
1:56:58the tools to be able to gain access and
1:57:00again this is the role of the school. So
1:57:03perhaps the administrators can start
1:57:04thinking about their their role in in uh
1:57:08providing or enhancing the the or or
1:57:11reducing the the income distribution of
1:57:15of uh through through their through
1:57:18their support uh services to these
1:57:20marginalized students.
1:57:24>> Thank you uh so much uh profess uh Dr.
1:57:27Francis now for your insight. uh clearly
1:57:29AI is not um only transforming
1:57:32industries but also our academic
1:57:34institutions know affecting the skills
1:57:37uh needed uh from both the academ.
1:57:40So this really shows uh the urgent need
1:57:43to strengthen our capacity building
1:57:45efforts moving forward. So Siguro we can
1:57:47hear um additional insights from Chelsea
1:57:50on this.
1:57:52>> Thank you Marvin. So for me at the
1:57:54institutional level uh education
1:57:58administrators can promulgate AI ethical
1:58:01guidelines on how AI should be
1:58:03implemented within their institutions
1:58:06which must be anchored on a human
1:58:08rightsbased approach and core principles
1:58:11such as ethical and responsible use and
1:58:13transparency. So that this guiding
1:58:16framework will enable educators to
1:58:20conduct and uh develop their competency
1:58:24assessments and as well as their
1:58:26curriculum pattern towards the use of AI
1:58:29in the classroom. And then at the
1:58:31educator level, let me highlight the
1:58:33critical role of these teachers because
1:58:36they have the responsibility to guide
1:58:39their students on how AI should be used
1:58:42as a complement in their need. So this
1:58:45should not be used as a teacher where
1:58:47they could provide uh they could ask the
1:58:49AI and the AI would generate certain
1:58:52information because they are still at
1:58:54the learning stage and they must have a
1:58:56good grasp of foundation. they might be
1:58:59prone to hallucination or biases
1:59:01provided by that generative AI. So just
1:59:04to provide a concrete example supposedly
1:59:08that I am still in my elementary or high
1:59:10school uh high school level and then I
1:59:12asked the AI what is 1 + 1 and then the
1:59:15AI provided me with an answer of three.
1:59:18So for me I might believe that that that
1:59:21three is 1 + 1. But when you're an
1:59:24educator, when you already have the
1:59:25foundation, you know that that is not
1:59:28wrong. But these students are not yet
1:59:30trained or they not um they do not have
1:59:33a good grasp of the fa basics and um
1:59:36foundational concepts. Hence, they would
1:59:38be prone to this kind of misinformation
1:59:40or they would not they are not trained
1:59:42to detect these biases provided by the
1:59:45generative AI. instructors can at least
1:59:48use um artificial intelligence or
1:59:50introduce what AI tools that would be
1:59:53used by students so that these
1:59:56instructor can lead um these students on
1:59:59how AI could be ethically and
2:00:01responsibly used and I'd also also like
2:00:04to highlight the role of the education
2:00:07administrators that AI should also be
2:00:11um increased through providing trainings
2:00:14or capacity building programs. on AI
2:00:17literacy among their academic and
2:00:20non-academic staff because um it's a
2:00:23really an institutional approach on AI
2:00:26adoption. It should not be at the
2:00:28individual level or at least at the
2:00:30classroom level. The institution has a
2:00:33big role to play if they really want to
2:00:35usher their students also to the future
2:00:38of learning where AI will be used as a
2:00:40compliment and this would significantly
2:00:43affect their um entry in the labor
2:00:45market if they are not guided on how AI
2:00:48should be used ethically and
2:00:50responsibly. So let me just provide a
2:00:52use case in research. So it is very
2:00:55common that certain students may just
2:00:58use or copy paste the outputs generated
2:01:01by AI even copy pasting that um AI said
2:01:05Chad GT said like that and that is very
2:01:08easy to detect by the instructors or the
2:01:11uh professors and um they are not really
2:01:15restricting the use of AI and as a guard
2:01:18rail certain institutions academic
2:01:21institutions are implementing that once
2:01:25a student is in their thesis stage, they
2:01:27should provide a declaration of AI boost
2:01:30which we also use in our AI study. We
2:01:34had this uh initiative that since we
2:01:36also used AI in the study, we declared
2:01:38what AI tool we used, how we use that AI
2:01:42and what human intervention that we
2:01:45provided because it promotes
2:01:47transparency and responsible use and
2:01:49ultimately it means that we as humans
2:01:52and individuals still exercise our human
2:01:55judgment on the outputs generated by AI.
2:01:58So I think this also um ends my answer
2:02:02and I hope our uh academ
2:02:07um attendees from educational
2:02:09institutions would at least consider um
2:02:12the answers that I provided. Thank you.
2:02:16>> Thank you Chelsea. And uh speaking of
2:02:18academic institutions, no we uh in the
2:02:20government also help uh promote
2:02:22evidence-based policym particularly as
2:02:25we navigate the impacts of artificial
2:02:27intelligence. That's why um open library
2:02:30access platforms um such as SERPI and
2:02:33Institute for Labor Studies are uh very
2:02:36important because it connects the academ
2:02:39uh research institutions making credible
2:02:42research uh more accessible and useful
2:02:44for policy and decision making. After
2:02:47all no um it's really about making sure
2:02:50uh our policies are informed, relevant
2:02:52and responsive to today's challenges. So
2:02:55let's move on to our uh next uh
2:02:59question. Um this is from Facebook no
2:03:03from uh Miss Mendoza and this is
2:03:05specifically directed to um Sir Francis.
2:03:08The question is um there is an inverse
2:03:12correlation between LGU's income and AI
2:03:15readiness. Does this mean we should tap
2:03:18national budget to strengthen lower
2:03:20income LGU's capacity to be AI ready?
2:03:24And uh another related question from
2:03:26Alexander Machinano of CPD Karaga
2:03:29region. If AI readiness is the new
2:03:32infrastructure, does your data suggest
2:03:35that investments may be premature for
2:03:38LGUs that still face income and
2:03:41connectivity gaps
2:03:46that that means they were really paying
2:03:47attention to my presentation. So those
2:03:50are really very good questions. No. So
2:03:52but first uh let me just clarify there's
2:03:54actually a the relationship is not um
2:03:59inverse no so it's actually if you will
2:04:01look at it if you're the richer you are
2:04:04your the more you are ready but um I
2:04:06think uh the the confusion might come
2:04:09because of the ranking because if you're
2:04:11a first class uh an LGU that means you
2:04:14are richer and then that means you are
2:04:16more ready but the first is actually a
2:04:19lower number so that's where The
2:04:21confusion is it becomes a negative or an
2:04:24inverse uh relationship or the ranking
2:04:27is uh actually inverse but but what this
2:04:30really means is if you are richer you're
2:04:32actually um have more uh uh AI you're
2:04:36more uh AI ready no so that's uh
2:04:38actually what what that means no so what
2:04:41uh what we are also suggesting is that
2:04:43for the fourth fifth and sixth class
2:04:45municipalities remember one of the uh
2:04:47recommendations that we provided is to
2:04:49actually have this fund that can
2:04:51actually help support the the the
2:04:54adoption needs of the fourth, fifth and
2:04:56sixth class municipalities. And then for
2:04:58the first and second class
2:05:00municipalities or or first and second
2:05:02class uh LGUs the the range is actually
2:05:05dependent on how they are actually
2:05:07performing if they actually do still
2:05:09need it. So in a way yes there is
2:05:12actually a um a need to really support
2:05:15the um especially those in the GA areas.
2:05:19No. So those who are really um lacking
2:05:22the the infrastructure to utilize uh
2:05:25these uh technologies um
2:05:29okay um yeah so so
2:05:32yes um there is a need for for national
2:05:35budget to to become more redistributive.
2:05:38So if there's already very strong
2:05:40performance in in the richer uh areas
2:05:44especially in NCR there is really a need
2:05:46to redistribute and to focus attention
2:05:48for areas so that they would be able to
2:05:50catch up. So if AI readiness is the new
2:05:53infrastructure, does your data suggest
2:05:55that investments may be premature uh for
2:05:58um actually um
2:06:02what we what we actually saw is that
2:06:04there's a need to do these to do um to
2:06:08implement strategies at a multi-pronged
2:06:11level. No. So you don't really
2:06:12prioritize one and then um uh just um
2:06:16wait until you completely finish this
2:06:18before you move on to the next. is
2:06:20actually implementing strategies um
2:06:23simultaneously because um technology
2:06:27changes so fast. No. So while we are
2:06:29looking at implementing digital commerce
2:06:32um digital services in in the LGU, we
2:06:36should also already be thinking about
2:06:38utilizing the information gathered from
2:06:40here in order to come up with better
2:06:42services and better um uh how these can
2:06:46actually feed in as data for um AI or
2:06:50for um ease of providing better services
2:06:53to to to
2:06:56the constituents. So for example, it
2:06:58allows us to identify areas where
2:07:01there's a concentration of uh people uh
2:07:04who are vulnerable to disaster or um it
2:07:07allows us to identify areas of of uh the
2:07:10need for um the phased um provision of
2:07:15uh social services because we would be
2:07:17able to identify um whether which months
2:07:20of of the year do people really need
2:07:23more uh income. No. So, so
2:07:27but these are just um information these
2:07:30would just remain as information if they
2:07:32don't are not really um processed well
2:07:36and u made available for um large large
2:07:40language models or for AI applications
2:07:42to make in um insights and come up with
2:07:45insights from these data. So that's why
2:07:48it's um while infra we're focusing on
2:07:50infrastructure it's also important to
2:07:52look at the the skills of the people in
2:07:55the LGUs are they able to make um to
2:07:59utilize these uh information as well to
2:08:01come up with these kinds of insights and
2:08:04um again corlary to that no so you have
2:08:07all all these prongs you should also be
2:08:09a the LGU should also be able to come up
2:08:12with an enabling framework that allows
2:08:15the the LGU do to extract information
2:08:18without violating also the privacy uh in
2:08:22of of of these uh um uh uh of their
2:08:26constituents. No, so of the data that
2:08:28they are extracting. So again um the the
2:08:31information should be or or the
2:08:33implementation should be um on a
2:08:35multi-prong level implementing them uh
2:08:38simultaneously.
2:08:40Thank you.
2:08:41>> Thank you so much uh Dr. Francis for
2:08:43your insight. No, it's very clear that
2:08:45uh AI readiness across uh LGUs uh is
2:08:48still an event especially uh in the
2:08:50far-flung area. So this uh also
2:08:53highlights the ongoing challenge of uh
2:08:55the digital divide. Uh moving forward
2:08:58know there's really a need to strengthen
2:09:00support particularly in terms of uh
2:09:02infrastructure
2:09:03uh skills and more equitable resource
2:09:06distribution so that LGUs can uh keep
2:09:08pace with the digital transformation. So
2:09:12um we move on to the next question. This
2:09:15is from uh Rico Marcado of Right Lens
2:09:18AI. This question is directed to
2:09:20Chelsea. Uh the question is based on
2:09:23your findings, do you see governance as
2:09:26something that should uh precede broader
2:09:29AI adoption or as something that
2:09:30typically follows once capability
2:09:33improves?
2:09:38For AI governance, I believe it should
2:09:40be at least flexible because AI
2:09:44technology moves us at a faster pace. So
2:09:49um regulation um should also be as
2:09:52flexible or we can also try to see how
2:09:54other countries are uh using um AI on
2:09:59how to improve their services the
2:10:02regional level. If I'm not mistaken, um
2:10:04Vietnam is the first Asan country to uh
2:10:08enact an AI law while uh other
2:10:10institutions or here in the government,
2:10:13we are also looking on how the EU um is
2:10:18has enacted the AI law because they uh
2:10:21at least proide a riskbased approach on
2:10:24regulating AI. So just to say that our
2:10:28AI governance is still at the
2:10:31formulation stages. We are looking into
2:10:33how other countries are implementing
2:10:35their um uh regulating AI. But suffice
2:10:39to say that our our governance system
2:10:43should not put guard guardrails on AI
2:10:45because it could also impact our
2:10:48economic development. Our enterprises
2:10:50should have at least the flexibility.
2:10:53Our governance system should have at
2:10:55least the flexibility on how AI
2:10:57governance should be implemented here
2:11:00that incorporates a human based approach
2:11:02and also a risk based approach in
2:11:05managing different levels of AI systems.
2:11:10>> Thank you uh so much uh Chelsea. Okay.
2:11:12So, Indino there's also um so much we
2:11:14can learn from um other experiences from
2:11:17other countries as we refine our own AI
2:11:20approach to AI governance here in the
2:11:23Philippines. So, uh for our next
2:11:25question, this is for Professor Alexis
2:11:28again. Um this is from Alessandro Casana
2:11:32uh from Zoom. Uh the question is what do
2:11:35you think is the timeline for some of
2:11:38the societal changes you mentioned in
2:11:40your presentation?
2:11:44>> Thank you. I mean we cannot really pin
2:11:47down the time but we can say with
2:11:50certainty that before changes would take
2:11:52thousands of years then they would take
2:11:55centuries then they would take decades.
2:11:58I think the changes we see now will take
2:12:01uh place in the next two three years
2:12:06because once everybody switches to using
2:12:08LLM models to get tasks done they will
2:12:12do it in the next weeks or months
2:12:14actually. So we'll see I I think the
2:12:16changes will be in the next year or two.
2:12:19Those that manage to learn how to
2:12:21leverage AI to do stuff to create stuff
2:12:25will benefit. the rest will suffer. Now
2:12:28it doesn't mean for everybody it will be
2:12:29in two three years. It depends what
2:12:31where you live, what type of job you do,
2:12:33how you engage with AI. But I don't
2:12:36expect decades. I expect this to take
2:12:38you know a few years.
2:12:44>> Thank you uh so much uh professor
2:12:46Alexis. So um we have another question
2:12:49from Zoom and this is directed to all
2:12:51our uh presenters. Uh let's start with
2:12:54uh Dr. Francis then Prof. Alexis and
2:12:57then Chelsea. So the question is how can
2:13:00uh private sector employees better
2:13:02prepare for the future of work
2:13:04especially in global industries and what
2:13:07government programs or trainings are
2:13:09available to support AI readiness. Let's
2:13:14start with Dr. Vis.
2:13:18Okay. So um I think
2:13:23Okay. So I I think privately no as
2:13:25individuals we are already exposed to
2:13:30the potential of uh the use of AI. No.
2:13:33So even if the institution we are in or
2:13:36does not uh have a formal use AI use
2:13:40policy, we are actually using um AI to
2:13:43do some of the tasks that often has
2:13:46eaten a lot of our time. No. So
2:13:49processing uh emails or creating
2:13:52responses to emails uh checking grammar,
2:13:56editing some of these things and even uh
2:13:59basic searches are also allowing us to
2:14:02do uh so the search cost has really um
2:14:06reduced on the time uh that's being
2:14:08spent on on that. No. So search cost has
2:14:11al also gone down. So uh as uh as
2:14:15employees no so what we are seeing is
2:14:17that some of us who are are also already
2:14:21being um made more aware of the a of the
2:14:25potential and the use of AI in our main
2:14:28daily tasks are actually improving. So
2:14:32um then the question is how what else
2:14:34can we do no? to um try to come up with
2:14:38so with the time that you save with the
2:14:41cost and efficiency that you gain. So
2:14:43what else can you do with with your your
2:14:45time? So are you being more productive
2:14:47with with the time that um you get and
2:14:50can are you also exploring other avenues
2:14:53where your your um skills can also be
2:14:56used. So that's actually what um uh what
2:15:00uh the employees can actually look into
2:15:03look beyond. No. So um
2:15:07and then also look at um what more can
2:15:10you do? No. So it's it's actually
2:15:12looking beyond what the
2:15:15what your position in the in the team is
2:15:20actually providing you. So what are the
2:15:21opportunities in the the the
2:15:25um company is providing you because the
2:15:28the technology is there and um as
2:15:31individuals you're also trying to
2:15:33familiarize yourself with the use of
2:15:35technology but then um
2:15:39what else what else can the tool do for
2:15:41you and for your company and and that's
2:15:44the thing that where where you need to
2:15:46to look at and then try to crit also
2:15:49critically look at yourself. No. So if
2:15:51you are um if you have seen that you
2:15:54have gained uh time, you have gained um
2:15:57efficiency becoming more productive,
2:16:00try to see what else can you create and
2:16:03look go to look at the government and
2:16:06skills that um skills training that's
2:16:09being provided so that you would be able
2:16:11to utilize this uh uh these tools for
2:16:15for the creation of other things. No. So
2:16:17I I think that's uh um what you can do
2:16:21as as um an employee
2:16:24and in these businesses.
2:16:27>> Thank you.
2:16:28>> Thank you uh Dr. Francis for your
2:16:30insights. Uh perhaps maybe we can get uh
2:16:33the thoughts of um professor Alexis on
2:16:37this from a global perspective.
2:16:40>> Okay, sure. Uh so I think the first part
2:16:42is looking at the work that Francis and
2:16:45Chelsea are doing at their institutions,
2:16:47collecting data, having surveys gives a
2:16:49lot of information on where the
2:16:51opportunities may be, the challenges,
2:16:54uh what tasks are affected and so on. So
2:16:57you can see in your own domain what
2:16:59other people thinks. But individually, I
2:17:02think regardless of where you work and
2:17:04what tasks you do, you shouldn't rely on
2:17:07the government to hold you by the hand
2:17:09and bring you to the right training. you
2:17:10know there's a resource constraint the
2:17:12government is always behind the curve
2:17:14and it's not their problem this thing is
2:17:16moving very very fast so you you you
2:17:19have to you know but you can do it on
2:17:20your own for example you can take many
2:17:23short courses online we all have free
2:17:25time we spend time watching rails you
2:17:28can take free courses online all you
2:17:30need is internet connection you can get
2:17:32free access to some of the LLM models
2:17:36and and you can ask how do I even start
2:17:40learning learning about it and they will
2:17:41tell you. You can say I work in this
2:17:43job. I have to put these documents
2:17:46together and I'm trying to think whether
2:17:48AI is affecting me. Start a conversation
2:17:51and you will see how fast you will
2:17:53advance that conversation. You have
2:17:56somebody to talk to. Just try that. Just
2:17:58get your hands wet. Just try it and you
2:18:00will see that you learn. uh not only
2:18:03will you find ways to improve the
2:18:06delivery of the tasks you have, it's
2:18:08like you know trying to add numbers and
2:18:10you don't know how to use a calculator
2:18:12once somebody showed you it became very
2:18:15important. Then somebody showed you how
2:18:17to use Excel. You said wow how could I
2:18:20even do my work without knowing Excel
2:18:22and then somebody showed you Tableau
2:18:25PowerBI or now AI. So discover these
2:18:28things because it it improves how you do
2:18:30tasks and increases your mobility, your
2:18:33ability to move from one job to the
2:18:35other. So that's a matter of job
2:18:37security and safety. But the other thing
2:18:39is by learning these tools, you may
2:18:41discover talents you didn't know you
2:18:43have. Like you may discover that you
2:18:45know I like to build things in specific
2:18:47domains and I just never thought about
2:18:49it because I never had the chance. So
2:18:51have fun with it. explore, but don't
2:18:53don't sit back and wait for the
2:18:55government because it will be too late
2:18:56and it's not the government's fault.
2:18:58It's not that the government doesn't
2:18:59it's the nature of this beast. It's
2:19:01moving very very fast.
2:19:05>> Thank you so much. Uh, Professor Alexis,
2:19:07uh, maybe we can also hear from Chelsea,
2:19:10you know, uh, coming from the government
2:19:12side, um, especially on the capacity
2:19:14building programs and the initiatives in
2:19:16place to support AI readiness.
2:19:20Thank you Marvin. I think first
2:19:22regardless whether you are a private
2:19:25sector employee or coming from the
2:19:28public sector I believe that all
2:19:31employee should embrace an attitude for
2:19:34lifelong learning because our current
2:19:36discussion is frame of artificial
2:19:38intelligence or generative AI. It could
2:19:41look different in 10 years time. It
2:19:43could be another emerging trend by 2040
2:19:47or by 2050. So to really futureproof an
2:19:50employee, you should adopt this type of
2:19:53attitude so that you will be keeping up
2:19:55with these trends in the labor market,
2:19:57these shocks in the labor market. And if
2:20:00you are able to at least cultivate this
2:20:02attitude towards lifelong learning, you
2:20:04will be self-sufficient and you will
2:20:07teach yourself on how to um use or be
2:20:11updated with these type of trends. So
2:20:13right now we are in the world of work
2:20:16where it is now defined by artificial
2:20:18intelligence. So for me since um I am
2:20:21still developing at least this attitude
2:20:23of lifelong learning on my own level I
2:20:26try to at least see on how AI is being
2:20:28used for my daily tasks. I I try to look
2:20:32for how um AI is being used in research.
2:20:35really try to devote time on what AI
2:20:38websites that is specific for research
2:20:40that I can um use for conducting my
2:20:44research and if you are still not aware
2:20:47of how AI is being used um you should
2:20:50also know that you don't have to switch
2:20:52industries and I think that's what sir
2:20:54Alexis is also emphasizing you can
2:20:56capitalize on your industry knowledge
2:20:59and you can look for specific AI
2:21:01training so that you can be a data
2:21:03analyst a data engineer
2:21:05or at least a machine learning
2:21:08specialist or at least an AI trainer
2:21:10because it is already your expertise. If
2:21:13I am from the banking and finance and
2:21:15then I try to upskill with AI related um
2:21:18skills, I am now an AI skill. I am now
2:21:22an AI expert on the banking and finance
2:21:24industry. You don't have to at least
2:21:26switch industries because of artificial
2:21:29intelligence. build upon your existing
2:21:32skills and knowledge and that will be um
2:21:34your niche market in the industry. And
2:21:38for the government programs, the ICT is
2:21:40providing free training programs um in
2:21:43their regional offices or in their head
2:21:47offices. You just have to look for their
2:21:49Facebook pages on or the regional
2:21:52officees Facebook page. Tesla is also
2:21:56providing training regulations or
2:21:58competency standards for data analytics
2:22:02and other um AI related skills such as
2:22:06image labeling. And then you can also
2:22:08look for other trainings on other
2:22:11government institutions such as the
2:22:13development academy of the Philippines
2:22:16because back then they were they are
2:22:17implementing the Sparta training
2:22:19program. And if these government courses
2:22:23would not be aligned to your skill or to
2:22:26your time or like your resources certain
2:22:30um websites are also providing free
2:22:34training. So this learning management
2:22:36platforms are easily searchable in the
2:22:39internet and they provide free courses
2:22:43on data analytics and science just to
2:22:46stop jump start your um AI journey
2:22:51that's all Marvin
2:22:54>> thank you so much Chelsea know so this
2:22:56really reminds the us that uh we must
2:22:58keep developing our skills uh taking
2:23:01every opportunity um to learn adapt and
2:23:04evolve especially the government
2:23:06trainings um provided by our um
2:23:09government institutions. So um we
2:23:12received an overwhelming number of
2:23:14questions from Zoom and Facebook but uh
2:23:17due to our limited time um please allow
2:23:20me to read the final question. So this
2:23:23is directed to all our presenters. Uh
2:23:25let's start with Professor Alexis uh to
2:23:28be followed by Dr. Francis and then
2:23:30Chelsea. So this is uh again from Miss
2:23:33Mendoza from the House of
2:23:35Representatives. The question is uh what
2:23:38higher level competencies should
2:23:40Filipinos develop in order to ensure
2:23:44their market ability in the labor market
2:23:46in the age of AI.
2:23:49Let's start with um Dr. Francis. I I
2:23:52I'll start with Prof. Alexis and then
2:23:54Dr. Francis and then Chelse.
2:23:58>> Okay. Thank you. So um I want to make a
2:24:01comment. It's relevant to this question
2:24:02and the one before and I think it's the
2:24:04most important comment I'll make in this
2:24:06presentation. So when I talk to these
2:24:08VCs that invest in California and they
2:24:11look at 1,000
2:24:13pitches every every year where to to
2:24:16invest their money and one thing they
2:24:18tell me is they like founders that are
2:24:20not tech people. They like find founders
2:24:24that are 40 50 years old. they work in
2:24:26some sector in the economy and they
2:24:28identify the real problem
2:24:31and then you know you can figure out the
2:24:33tools AI will actually tell you what
2:24:34tools you need to solve a problem the
2:24:36problem is the people that know AI and
2:24:38they know technology they never worked
2:24:40in real life and they don't know real
2:24:42problems to solve and what you will see
2:24:44more and more these days is people from
2:24:47the industry people solving very
2:24:50difficult problems in very se various
2:24:52sectors they realize the inefficiency
2:24:55They realize that what they do is very
2:24:57inefficient and technology can improve
2:24:58it. Maybe they didn't know how to solve
2:25:00it. They don't need to know how to solve
2:25:02it because AI will tell them. But they
2:25:04understand the business. Somebody from
2:25:06the outside doesn't understand the
2:25:08business. They don't understand the
2:25:10problems. But now that has changed that
2:25:13creates a lot of opportunities.
2:25:17So the first competence is you know try
2:25:19to understand your domain very well.
2:25:21That's your your stronger skill. The
2:25:24second one, everything is about data.
2:25:27You know, data visualization for me is
2:25:29very important. Every transaction we
2:25:30make, every interaction generates data.
2:25:33We need to understand data. And finally,
2:25:35play with AI because I may decide that I
2:25:39want to renovate the room behind me and
2:25:41I don't know modeling. I don't know how
2:25:43to go on the computer and do renderings
2:25:45and then see which one is better or not.
2:25:47But if if Marvin tells me, you know,
2:25:49there is this tool, you go in, you put
2:25:52some photos of the space, some photos
2:25:54that you like. You explain why you like
2:25:56it and give some directions what you
2:25:58like, what you don't like. It will
2:25:59create space for you. And if I'm able to
2:26:02communicate that in 5 seconds, I'm able
2:26:05to redesign my space. I have an idea. So
2:26:08then it's really about how you
2:26:09communicate, how you articulate, but
2:26:11somebody has to tell you that that tool
2:26:13exists. So a very important competency
2:26:15is just to know what's out there even
2:26:18with data visualization not that you
2:26:20have to do coding but to know what to
2:26:22ask AI. Thank you.
2:26:26>> Thank you uh professor Alexis maybe we
2:26:29can hear the insights of uh Dr. Francis.
2:26:33>> Thank you. That's actually a very good
2:26:34question and uh uh I it it brings me
2:26:38back to a classmate of mine when I was
2:26:40in my my uh taking my masters in in in
2:26:44Atena when um uh he he actually when
2:26:49when I asked I posed him a question so a
2:26:52short story I posed him a math question
2:26:54and then he said that you know that's a
2:26:57hard question and I can only give you my
2:27:00answer in two things no so it's either
2:27:03either I know the answer or I know where
2:27:06to get the answer. So that's there's the
2:27:08only two things that he knows. He said I
2:27:10know where uh I it's either I know the
2:27:12answer or I know where to get the
2:27:14answer. And I think it's something
2:27:15that's very important for all of us here
2:27:18who are now facing this challenge or
2:27:20this major question of what how will AI
2:27:23affect all of us? No. So it's
2:27:26it's very difficult for us to already
2:27:28know the answer. But perhaps we can
2:27:31utilize this tool that's already in
2:27:33front of us to
2:27:35use it to get some answers. No, to we
2:27:38don't know what the answer might be, but
2:27:41I I think it can be used to get uh an
2:27:44answer. So, and if and I think that's
2:27:49that's a good starting point for for all
2:27:51of us uh here who are asking so what
2:27:53competencies do I need to know uh in
2:27:56especially for uh where I am? I think
2:27:58one thing is also first is asking the
2:28:01right questions. No. So if you are in in
2:28:05the development space for like all of us
2:28:08it's important to ask um to have that
2:28:11curiosity of uh asking um why such
2:28:14things are or why are people behaving in
2:28:18a certain fashion. So and and
2:28:21um that's also a good starting point for
2:28:24AI. No, because in AI you ask you you
2:28:26you you feed them with questions. No, so
2:28:30you and the questions will only come
2:28:32from you. No, they from your own cur
2:28:34curiosity from your own um experiences
2:28:38in the in the environment and in the
2:28:40circle where where you are. So um
2:28:43formulating that kind of question is
2:28:46something that's very important that
2:28:47that competency will will uh help you um
2:28:51identify and extract the the answers
2:28:54that you you are looking for. I I think
2:28:57it's al something very important also is
2:28:59um
2:29:01empathy um the being part of knowing
2:29:07that you are also not just by yourself
2:29:10and being part of a bigger environment
2:29:12that where where you are no so that
2:29:15builds on on curiosity and and trying to
2:29:19um feel what the others are feeling. So
2:29:23again asking the right questions why
2:29:25certain things are and why um um um
2:29:30these people are experiencing that kind
2:29:32of experience even if you are not
2:29:34experiencing it. No. So why are people
2:29:37becoming more burdened by disaster or
2:29:40why are um middle class being asked to
2:29:43pay more for certain things? No. So why
2:29:46are taxes being imposed on all of us? um
2:29:51and why are certain beneficiaries
2:29:54being on the receiving end of certain
2:29:57beneficiaries. So understanding and
2:29:59asking those kinds of questions and
2:30:01being critical about the answer that AI
2:30:04will give you. So the next competency
2:30:06would be again critical thinking. So not
2:30:09just receiving the answers uh that I I
2:30:12would um on face value but really trying
2:30:15to identify um what the um how this can
2:30:19be improved and whether these are
2:30:21actually based on on sound uh data and
2:30:26then finally I think this is a a
2:30:28competency that we don't really um see
2:30:31often but risk takingaking uh being um
2:30:35more uh
2:30:38averse risk averse to technology
2:30:40something that we we would um uh want
2:30:44it's it's something that we need to
2:30:46really expose ourselves no so when
2:30:48people say that oh that's AI I'm afraid
2:30:49of that we really try should try to
2:30:52discourage them um but rather equip them
2:30:55with the capacity to to understand what
2:30:58it can do what it can't do and what type
2:31:00of information you should feed into that
2:31:03um um technology know
2:31:06um and use that technology, take the
2:31:09risk to use that technology for for your
2:31:13improvement for your other skills. So
2:31:15those are some of the competencies I
2:31:16think that um would would be important.
2:31:21>> Thank you so much um Dr. Francis. So
2:31:24indeed we know it really comes down to
2:31:26asking the right questions, critical
2:31:28thinking, having empathy especially in
2:31:31evidence-based policy making so we can
2:31:34stay curious um adapt to change and
2:31:36ensure um that uh this transition
2:31:39remains human centered. So maybe we can
2:31:41hear also from uh Chelsea from the
2:31:44governor perspective.
2:31:46>> Um thank you Marvin. Um I believe uh we
2:31:51should focus more on the soft skills of
2:31:54the um job seekers because nothing could
2:31:58really replace the soft skills. It is
2:32:00very different from a lot of people and
2:32:03we should also remember that training
2:32:06institutions and educational
2:32:08institutions are also collaborating with
2:32:11the industry to prepare the students for
2:32:13their technical skills. So basically
2:32:16everyone especially if you're belong
2:32:18into a specific industry have a leveled
2:32:21playing field because of the skills that
2:32:25you learn and the knowledge that you
2:32:26have learned in these uh school and
2:32:29training school and training
2:32:30institution. But what makes you
2:32:32different is your soft skills. Employers
2:32:35I think also based on a ILS study on Gen
2:32:38Z and the youth employers are really
2:32:40more looking on to the soft skills of
2:32:43the job seekers because people are
2:32:47attainable and if you are attainable
2:32:49with the skills you're um the other
2:32:52factor that they will look into is your
2:32:55soft skills. Are you uh do you have
2:32:57collaborative
2:32:59uh engagement with your teammate? Do you
2:33:01know how to communicate with your
2:33:03stakeholders? Can you solve problems on
2:33:06your own? Can you provide critical
2:33:08thinking? And also, can you exercise
2:33:11leadership and management your um lower
2:33:15staff? So, I believe soft skills should
2:33:17really be developed continually and we
2:33:21should have an attitude towards learning
2:33:23because the market is always changing.
2:33:26As I said earlier um the technology that
2:33:29we know now could be different by next
2:33:32um year even next year in the next 10
2:33:35years or by 2050. So as the market
2:33:38changes, as the industry changes, our
2:33:41way of future proofing ourselves from
2:33:44these um macro shocks is to develop our
2:33:48own soft skills and adopt an attitude
2:33:51for life lifelong learning. So that when
2:33:54we experience this macroeconomic shops
2:33:56or shops in the labor market, we know
2:33:58how to handle ourselves. We know how to
2:34:00train ourselves on the current trend. we
2:34:03know where to go to to look for scaling
2:34:05programs and that we are also well um
2:34:08adaptable to these changes in the
2:34:11workplace. So that um that's my answer
2:34:15Marvin. Thank you.
2:34:17>> Thank you uh Chelsea. So um as
2:34:20reiterated by Chelsea, you know, for job
2:34:22seekers, yes, technical skills can be
2:34:24learned, but it's really the soft
2:34:26skills, how you think, how you
2:34:28communicate, and how you adapt uh that
2:34:30will make you um competitive in the
2:34:33labor market. So um thank you very much
2:34:36um for your active participation to our
2:34:39participants in our open forum. So we
2:34:41will now uh conclude our discussion and
2:34:44move on to the closing messages. So let
2:34:46us call on our panelists uh one by one
2:34:49to share their very brief final
2:34:52reflections on today's discussion. So we
2:34:55will begin uh with uh Chelsea uh to be
2:34:58followed by Dr. Francis and then lastly
2:35:01professor Alexis. So Dr. Francis is also
2:35:04joined by his co-author um Sir
2:35:07Christopher on Zoom and we would also
2:35:09like to ask for his reflection after Dr.
2:35:12Francis. So um let's begin with uh the
2:35:15message of Chelsea.
2:35:18>> Thank you Marvin. First I'd like to
2:35:20thank our attendees because while um I
2:35:23was also waiting for my turn we are
2:35:26already raising questions that also uh
2:35:29deserve clarity from the presenters. And
2:35:31I think at my own level, it also caused
2:35:34a reflection on my part that we we
2:35:37should also learn how to set our um
2:35:42expectations on how AI is being used in
2:35:45different industry. So I I got that some
2:35:48of our attendees are from the government
2:35:50sector, some are from the education
2:35:52asking questions on how AI should be
2:35:54used. And for me this reflects that our
2:35:57attendees are really curious on how AI
2:36:00should be ethically and responsible uh
2:36:03how AI should be ethically and
2:36:05responsibly used within in their
2:36:07institutions and they are just looking
2:36:09for frameworks. they are looking on how
2:36:12it can be implemented and I think that
2:36:15is a good sign because we are not afraid
2:36:17of AI but we are willing to engage we
2:36:21are willing to embrace AI but we just
2:36:23don't know um where to look and how to
2:36:25look for it so I guess I hope that in
2:36:29the future policy uh discussions um our
2:36:33conversations will be also on an
2:36:36exchange basis we have inputs from the
2:36:39industry we have uh exchanges with the
2:36:42academ and also with the government so
2:36:44that our journey towards AI adoption
2:36:47could be of a polation approach that
2:36:50links labor supply institutions to our
2:36:54labor demand institutions. That's all.
2:36:59>> Thank you Chelsea for your message. Now
2:37:01let's hear from Dr. Francis and also um
2:37:05Sir Christopher who's also joining us
2:37:07via Zoom.
2:37:08Uh thank you very much. Uh again also
2:37:10thank you to the organizers, thank you
2:37:12to the PIDS uh for the invitation to
2:37:15present uh our paper. So the paper is
2:37:19actually looking at um the one of the
2:37:23smaller units of our government. No, so
2:37:26those that are really closer closest to
2:37:28us. No, the the the local government
2:37:30units and we realize that um uh we need
2:37:34to look at their readiness because it's
2:37:37it's either we just allow the technology
2:37:41to pass through us and be reactive to
2:37:44how it is being used or we can actually
2:37:47take advantage of the opportunities uh
2:37:50that is um being provided to us by
2:37:52through the use of these technologies.
2:37:55So we realized through the study that uh
2:37:57our LGUs are are not may not be you know
2:38:01all not all of them are ready some of
2:38:04them have a really major barriers but
2:38:07what we understand is that there's
2:38:09there's a way to build this readiness to
2:38:13address the challenges
2:38:15it can be through investment through the
2:38:17training of the LGU uh workers through
2:38:21uh good planning through deliberate uh
2:38:23well-defined policy See, so these things
2:38:26can actually be addressed and can
2:38:28actually be done by the LGU um LGUs
2:38:32themselves. So what what is important is
2:38:35we ensure that we embrace uh the use of
2:38:38AI so that no LGU or no local government
2:38:42no one is left behind. So the the the
2:38:48success of the the the Philippines is
2:38:50actually um Philippine uh government
2:38:54Philippine public service can actually
2:38:56be supported and enhanced through the
2:38:59use of technology such as AI that is
2:39:01actually human- centered and um um
2:39:06through through an ethical and people
2:39:08centered use of of AI.
2:39:12>> Thank you uh Dr. Francis. Um, let's hear
2:39:15po from uh your co-author uh Sir
2:39:18Christopher.
2:39:20>> Hi. Hi. Good morning. Thank you so much
2:39:22um Marvin and thank you for the
2:39:24organizers for inviting me as well in
2:39:26this um event. So I'll just keep it
2:39:30short. So, so basically um today's
2:39:33presentations aside from our
2:39:34presentation um it they paint a
2:39:38consistent picture that artificial
2:39:40intelligence holds enormous promise um
2:39:44across the world and including here in
2:39:46the Philippines for public services for
2:39:49workers. Uh but that promise is not uh
2:39:54will not be realized automatically. So
2:39:56it really requires um deliberate
2:39:59investment in invest in infrastructure
2:40:01and skills and also clear governance
2:40:04frameworks and policies
2:40:06so that to ensure that no Filipino
2:40:08worker no community no LG is left
2:40:11behind. So thank you.
2:40:13>> Thank you so much uh sir uh Christopher
2:40:16and finally let's hear the final message
2:40:19of professor Alexis.
2:40:22>> Thank you. We already shared a lot of
2:40:24insights. I think I'll make three
2:40:26comments. The first one is there is
2:40:29change coming you know we cannot avoid
2:40:30it. The change is actually here. The
2:40:33second thing is while the change is here
2:40:35in ways we cannot understand. We
2:40:37shouldn't be afraid.
2:40:40We need to understand there is change
2:40:42there has always been change. Maybe the
2:40:44rate at which this change is coming is
2:40:47not what we expected but we shouldn't be
2:40:49afraid. We should try to figure out how
2:40:52we can um benefit from our own side. And
2:40:54we talked a little bit about that. And
2:40:56the third comment is, you know, we all
2:40:59need to come together in forums like
2:41:01this, in research, in drafting policies.
2:41:03It cannot be somebody behind an office
2:41:05writing policies and educational reform
2:41:08programs. We all need to come together.
2:41:10But also as individuals, it's up to us
2:41:13not to be afraid and do nothing about
2:41:15it, just to explore. It's like the first
2:41:18time you saw a car on the street, you
2:41:20didn't know what it was. Maybe you felt
2:41:22overwhelmed, but really you just have to
2:41:23make sure you don't get step over by the
2:41:25car. Eventually, you learn how to drive
2:41:27the car, how to improve your life. You
2:41:29don't need to worry about how to build
2:41:31cars. You don't need to worry how to fix
2:41:32it. If it breaks, somebody else will do
2:41:34it for you. Just how to coexist and use
2:41:37it to your benefit. Thank you very much.
2:41:41>> Thank you so much, Professor Alexis. And
2:41:43of course we'd uh like to thank again um
2:41:47uh sir Francis uh Chelsea our uh
2:41:50resource speakers who joined us today
2:41:53for um sharing their insights and uh uh
2:41:56for joining us in this webinar. Of
2:41:58course thank you very much to all our
2:42:00participants for your active engagement
2:42:02and for staying with us um until this
2:42:05part of the program. So, let's give
2:42:06ourselves a big uh virtual round of
2:42:09applause, claps, hearts, and conferries
2:42:12for everyone.
2:42:14Okay. So, now as we wrapped
2:42:26governance
2:42:27angi as your uh trusted source of
2:42:30evidence. So isaong free open access
2:42:33library policy research from over 75
2:42:37institutions. So whether you are a
2:42:39policy maker, a researcher or even a
2:42:42concerned citizen
2:42:45credible and evidence-based studies
2:42:47through surf and through this platform
2:42:50institute for labor studies
2:42:55researches and work and contribute
2:42:58evidence based policym on labor and
2:43:01employment issues. So zoom chat box po
2:43:04and Facebook comment section link. So
2:43:07feel free um to check it out and before
2:43:10we officially end the program announce
2:43:14raffle winners as a token of
2:43:16appreciation for your participation. So
2:43:20from zoom we have our first winner is uh
2:43:24Rosan Ray Rubio from the social housing
2:43:27finance corporation.
2:43:29Congratulations po.
2:43:32Our next winner n man po from Zoom we
2:43:34have Lan Espasio from the Philippine
2:43:38business for social progress.
2:43:40Congratulations.
2:43:42And uh from Facebook we have um Jane
2:43:46Makapagal Mulasa.
2:43:50So congratulations po winners. The Ps uh
2:43:55tech team will coordinate with you on
2:43:57how you can claim your prizes. So u
2:44:00please wait for the team's email. Please
2:44:02don't leave just yet no as we have a few
2:44:05reminders. Uh the speakers presentations
2:44:07will be uh available on the ps website
2:44:10through the link provided in the zoom
2:44:12chat and uh s Facebook comment section.
2:44:16So for our um zoom participants we
2:44:18encourage you to accomplish the feedback
2:44:21form provided in the chat box to help us
2:44:24further improve our program. So you will
2:44:26receive your ecertificate after
2:44:28submitting the completed form. For our
2:44:31Facebook live viewers, you may still
2:44:33receive an ecertificate through replay
2:44:36on demand. So simply scan uh the QR code
2:44:39or the clink uh the link provided in the
2:44:42comments and submit your response before
2:44:451000 p.m. today. So again uh the
2:44:47deadline is 10 p.m. today. Please keep
2:44:50in touch with us as well. Follow the Ps
2:44:54uh Dole ILS and SERP on social media and
2:44:58visit our websites for more updates and
2:45:00information. You may also visit uh
2:45:03www.pids
2:45:06uh.gov.ph/events
2:45:09for details on the PIDs um upcoming
2:45:12activities for May 2026 which is flashed
2:45:15on the screen. So once again, thank you
2:45:18to all our speakers, discussants and
2:45:20participants. Our key takeaway today is
2:45:23simple yet important. Just because we
2:45:26are not being replaced by AI today does
2:45:29not mean we are already secure
2:45:33transition stage
2:45:35changing jobs evolving skills
2:45:38requirements and new technologies.
2:45:45importance on proactive policies,
2:45:48stronger institutions to learn skills
2:45:51development. Yes, AI brings great
2:45:54opportunities but without the right
2:45:56support systems
2:46:00uh existing gaps. Hopefully today's
2:46:03discussion brings us one step closer
2:46:06inclusive and future ready Philippines.
2:46:09on behalf of Psy and Dolly ILS
2:46:14po for joining us today and uh have a
2:46:16great day. Thank you.
2:46:34>> [music]
2:46:46[music]
2:46:53[music]
2:47:08[music]
2:47:11>> Baby,
2:47:15[music]
2:47:24hey. [music]
2:47:31>> [music]
2:47:44[music]
2:47:50[music]
2:47:57[music]
2:48:03[music]

About this transcript

This page contains the full transcript of AI and the Future of Work: Implications for Skills, Jobs, and Public Sector Readiness by Philippine Institute for Development Studies, generated from the public captions YouTube serves with the video. The transcript has 23,115 words across 3,615 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.