YT YouTube2Text

YouTube2Text → Transcripts → Here's What I'd Learn Instead of AI Automation in 2027

Here's What I'd Learn Instead of AI Automation in 2027 — Transcript

by Nick Saraev · 5,552 words · 816 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00As somebody in the industry that's made
0:01it close to 10 million dollars through
0:02AI automation, I've noticed a lot of
0:04incentives perpetuating this idea that
0:06AI automation will last forever. Whether
0:09you are an automation specialist or
0:10somebody on social media or a company
0:12that sells AI SaaS apps or what have
0:14you. You know, you're afraid to talk
0:16negatively about the industry cuz you
0:17think it's going to hit you in the
0:18wallet. Um so maybe it will, but AI
0:21automation is not going to last forever.
0:23And then I'm making this video for two
0:24reasons. The first is that model
0:26progress on automation benchmarks like
0:29automation bench and GDP eval, uh two
0:31benchmarks that basically came out of
0:33nowhere and then in less than 6 months
0:35models are already scoring very very
0:36well on are showing us that AI agents
0:38can now autonomously automate a lot of
0:41the business without a human being in
0:42the loop at all. And so back in the good
0:44old days, you had to micromanage these
0:46agents to get them to do what you want.
0:47You had to point them at the API
0:48documentation and the endpoints and get
0:50them to, you know, do the connectors and
0:51stuff like that. Now it's literally
0:53just, "Hey brother, I got a task for
0:55you. Can you automate
0:56www.myhvacbusiness.com?"
0:59And it'll actually go and do a pretty
1:00good job. The second big thing is uh
1:03knowledge asymmetry, which is um you
1:05know, how back in the good old days,
1:07again, you had a lot of knowledge about
1:09AI and what it could do. And the people
1:10you were selling to had very little
1:12knowledge about AI and what it could do.
1:13And so, you know, the way that arbitrage
1:15works is you're compensated on the
1:16difference in knowledge between you and
1:18your customer. And so we had a very big
1:20difference, therefore we were
1:21compensated a lot. Well now that
1:23knowledge gap is it's closing. I mean
1:25it's not closed, you know, maybe instead
1:26of like here it's probably over here.
1:28But obviously there's shifts there,
1:30right? And if we keep that delta going,
1:31eventually we'll be at the point where
1:32most people know what most AI agents are
1:34capable of to approximately the same
1:36level that you or I know. So here's what
1:38I want to do. I want to talk a little
1:39bit about both of those. Really just
1:40drill into detail about what they mean,
1:43uh what strategic advantages they are
1:44unveiling for people like you or I, and
1:46then also where I see the industry going
1:48over the next year or so. I made a video
1:50very similar to this um I guess about a
1:52year ago now that talked about what I
1:53would learn instead of automation in
1:552026. It's one of my most watched
1:57videos. It was also met with an
1:58incredible amount of, you know,
1:58criticism and and honestly like dis-
2:01distaste because I think I just said the
2:03elephant in the room and people
2:04typically don't like when somebody pulls
2:05the wool out from over their eyes.
2:07Um so, I'm probably going to get some of
2:08the same down below in the comments. If
2:10I'm getting eviscerated right now, well,
2:11there you go.
2:12Why don't we start with the first, which
2:13is Automation Bench and GDP vow. So,
2:16both of these benchmarks basically went
2:17from models scoring zero on them to with
2:20Gemini Argon 4 um about 50 to 60%. If we
2:24linearly extrapolate this progress,
2:26every month they've added somewhere
2:27between 5 to 10% on them. That means
2:29that somewhere between December to maybe
2:31March, models will probably start
2:33scoring 100% on them. This doesn't mean
2:35that they're 100% the best, by the way,
2:38but it means that that benchmark, which
2:39people tried to make very difficult, has
2:41now been saturated. And it means that
2:43they'll probably have to make a few new
2:44benchmarks instead. They'll make like
2:45Automation Bench V2, which is supposed
2:47to be 100x harder.
2:49Well, you know what's going to happen
2:49with Automation Bench V2, ladies and
2:51gentlemen? It's going to get saturated,
2:52too. And instead of taking, you know, a
2:53whole year, it's probably going to take
2:55like two or three months cuz that's what
2:56we're seeing with other V2, V3, V4
2:58benchmarks. Human beings are losing the
3:00ability to accurately forecast how smart
3:02models are. And models know this. Like
3:04they're aware of this in their training
3:06environment. They know how to even game
3:07the system now, for Christ's sake. With
3:09models breaking into world governments,
3:11is getting really weird. So,
3:13anyway, that's Automation Bench. And,
3:15you know, if you think you can compete
3:16with an agent like that, you're out
3:18of luck, which kind of necessitates a
3:20change in how you operate, right? We saw
3:22a lot of this with like website builders
3:24back in the day. If you think about it,
3:25like back in the
3:26I keep saying back in the good old days,
3:28but back in the terrible old days, you
3:30know, if somebody wanted a website that
3:32required an extreme level of specialized
3:34knowledge around things like, you know,
3:36like coding and and hosting
3:38infrastructure and you know, how to like
3:40acquire a a DNS, you know, DNS
3:42resolution, setting up your I don't
3:44know, your your your TXT records and
3:46stuff like that, whatever. Back in the
3:48time, you had to do that a lot and it
3:50was worth paying somebody a lot of money
3:52in order to get them to do that for you.
3:53Nowadays, you can just go to an AI agent
3:56and you can say, "Hey, can you do this
3:56thing for me?"
3:57It's the same thing with like, you know,
4:00hosting platforms like Hostinger
4:02or you know, WordPress hosting platforms
4:04or I don't know, Wix, right?
4:06Squarespace. You can just jump on there
4:07and you can do most of it yourself. And
4:09that means that some people are going to
4:10be doing that for AI automation as well.
4:11Back in the day, you would have been
4:13looked at AI automation and been like,
4:14"Whoa, I don't know how the hell to
4:15automate my emails." But now you're
4:17going to be like, "Hmm, hey Claude, how
4:18do I automate my emails?" Oh, it's
4:20actually fairly easy. You just do
4:21whatever. Can you automate my emails?
4:22Yeah, for sure. Boom, token cost 10
4:24cents. It's not going to cost you
4:25$10,000 like it was just a year and a
4:28half ago.
4:29So, I say this not to alarm you, but I
4:31do say this to be realistic. It's
4:33happening. However, just because this is
4:35happening doesn't mean that, you know,
4:37you can't make money with it. To be
4:38clear, there are lots of website
4:39designers out there today that make
4:41bank.
4:42I know a few just off the top of my
4:43head. So, one of the guys that I work
4:44with for instance, this little website
4:45is worth $300,000. He's worked with like
4:48governments. He's worked with major,
4:49major brands and businesses. The reason
4:51why he can do this in an environment
4:53that is otherwise entirely commoditized
4:55is because he is extraordinary at one
4:57aspect of the business. And that aspect
4:59of the business is sales and marketing.
5:01You know, his websites realistically
5:02don't run on a super HTML, right? They
5:05don't run on ultra CSS. They run on the
5:07same damn HTML and CSS that all other
5:09websites run on, obviously.
5:11But he is able to, you know, sell the
5:13benefits and market his business to
5:15enough people that the law of numbers
5:17eventually works in his favor, right?
5:18Somebody is interested at the specific
5:20time and he strikes while the iron is
5:21hot that he can land these like large
5:23six-figure plus deals.
5:25Um, and so, that's probably where this
5:27market is going such that if you did
5:29want to, you know, compete in this
5:31industry, you would have to be truly
5:32extraordinary at something. Either you'd
5:34have to be extraordinary at sales and
5:35marketing like I talked about, or you'd
5:36have to be extraordinary at the actual
5:38product like, um, maybe like Brett from
5:40Design Joy who does similar website
5:41development stuff, uh, you know, up
5:43until, well, I guess he's still doing it
5:44right now, but but not as popularly uh,
5:46was doing. Or or something else. Now,
5:48most people aren't actually going to be
5:49able to compete like that because the
5:50reality is most people aren't
5:51extraordinary, you know? I'm not
5:53extraordinary at like 99.9% of what I
5:54do. It is unreasonable to expect to be
5:56extraordinary at, you know, everything
5:58to the point where I'd be able to
5:59compete with somebody that is
6:00extraordinary. That's just, you know,
6:01that's just how it is. Which means the
6:02expectations running that specific line
6:05of products sales have to go down.
6:07Second big thing is this knowledge
6:09asymmetry bit.
6:10Um you know, obviously people are
6:11learning about AI. Up until I think like
6:14a year and a half ago
6:15was an 80% of the world's population had
6:17not used a paid AI model. Maybe it was
6:21even more like 85 to 90% but like 80% of
6:24the world's population approximately had
6:26not used an AI model intelligent enough
6:28to like write a coherent email.
6:29Um they were rife with errors and it was
6:31just like really dumb and obviously like
6:33AI generated, right? Um today that
6:34number is probably closer to like 70%.
6:36So, we gained 10% in the last year or
6:38so.
6:39The reality is as we hit AGI, we're not
6:41just going to creep 10% 10% 10% every
6:43week or month. We're going to go like
6:4450% overnight. You know, there going to
6:46be demos the likes of which you all have
6:48never seen before circulating on, you
6:49know, CNN and Fox News showing what
6:51these agents are capable of. Every man,
6:53woman, child, dog, and cat on Earth will
6:56know the capabilities of AGI. And so, a
6:59lot of them will start thinking pretty
7:00quick like, "Whoa, wait a second. AI can
7:02actually do all that stuff." Hmm.
7:04Such that when you make a cool sexy
7:06offer like, "I'll actually whip you up
7:08a, you know, PDF showing all this stuff
7:10totally custom." You know, it's not
7:12exactly going to be anywhere near as
7:13impressive. And that is, you know, just
7:15how it just how it goes, right? The the
7:17floor is always rising in some way,
7:19shape, or form. And so, what that means
7:20is if you stay in the same place, okay,
7:23and the floor is always rising, you're
7:25not staying in the same place. What
7:26you're doing relative to the floor is
7:28you're actually going down. That's
7:29something that I feel like a lot of
7:30people don't really understand. The only
7:33way to actually maintain growth is
7:35actually growing. It's kind of like
7:36interest, right? If you aren't
7:38constantly, you know,
7:40um you know, accumulating 3 to 4%
7:42depends on where you are in the world,
7:43on an investment, you're not actually
7:45like making the same amount of money,
7:47you're not like leaving your your
7:48investments neutral. You're actually
7:49making less money because the interest
7:51and the background interest and
7:53inflation rate is slowly eating away at
7:54the buying power that you have. And so
7:56that's the way I've always thought about
7:56this business model. I've always thought
7:58about it as like, if you're not
7:59constantly innovating or improving on
8:01this business model in some way, shape,
8:02or form, you know, you are just staying
8:03in the same place, you're decreasing.
8:05And you would be well to, assuming this
8:07is something you want to do, like
8:08understand that. Just a good rule of
8:10life in general. Okay, so I've talked
8:12all day about, you know, scaring the
8:13out of you. What the hell is this
8:15business model even for then, you know,
8:16if AI agents are coming for everything
8:18and blah blah blah. Is it even worth
8:19getting into? Um, and I I got some news
8:22for you. Is AI automation even worth
8:24getting into in 2027?
8:27Well, interestingly, yes. Because
8:28everything that I've told you so far is
8:30not actually just happening to AI
8:32automation, it's happening to every
8:34other industry on Earth. The same thing
8:36that I talked about, you know,
8:38automation bench, GDP valve,
8:40uh, knowledge asymmetry window
8:41shrinking, it's happening to all
8:44white-collar knowledge workers. It's
8:46happening to accounting. It's happening
8:47to marketing systems. It's happening to
8:49sales. It's happening to MSPs. It's
8:51happening to IT. It's happening to
8:53virtually all internet infrastructure
8:55and architecture out there.
8:56It's happening in networking. It's
8:57happening to SaaS apps. It's happening
8:59to like literally design, anything that
9:01you could possibly think of. Everything
9:03that I just told you is occurring the
9:04same in automation as it is to that
9:06industry. And so it's not really like,
9:08"Oh my god, AI automation is going down
9:10the tubes. What the hell am I going to
9:11do?" It's like, "No, literally every
9:14industry on planet Earth right now,
9:15specifically the white-collar ones, uh,
9:18because those are the ones that are
9:19being first hit because of digitization,
9:20are being transformed by AI."
9:23And the people that are going to win in
9:24the next couple of years, presumably the
9:26next few decades as AI continues to
9:28reshape our economy,
9:30are not the ones that just sit around in
9:31the same place doing the same things,
9:33trying to sell the exact same offers, or
9:35whatever. They need to kind of get with
9:37the times and need to start sort of
9:38evolve. Think of like brokers, you know,
9:40like real estate brokers or
9:43I don't know, like car dealers or like
9:45mortgage brokers or how even a lot of
9:47law. These used to rely on like highly
9:50specialized knowledge
9:52in a domain that was very risk
9:54asymmetric where there's a lot of risk
9:56if things went wrong. If things went
9:57right, it was like, oh yeah, I guess
9:58it's kind of okay. Back in the day you
10:00used to have to pay these people tons of
10:02money in order to basically facilitate
10:03any transaction on planet Earth. And
10:06nowadays, most of the time you can just
10:07go on a website and directly purchase
10:09something. Most of the time these are
10:11the appendice appendix of like the
10:14modern economy. They're like slow,
10:16laborious things that are just between
10:18an Apple Pay payment and then you
10:19getting the thing that you want. You
10:21know, I'm looking for a commercial lease
10:22right now. And while I like the guy a
10:24lot, you know, it's it's quite difficult
10:25have to go through this intermediary
10:26process to ultimately do what I should
10:28be able to do at the click of a button.
10:30And so those are relics of, you know,
10:32like a past sort of environment just
10:34like I think most white-collar knowledge
10:36work today including AI automation will
10:37be relics of
10:39past environment looking forward. This
10:41is a wider shift I think we as a society
10:43have been under the process of for the
10:45better part of the last 20 or 30 years.
10:47Now, the difference is with our
10:49industry, and this is really the point
10:50to understand, which is why I decided to
10:52bury it 10 minutes into the video.
10:54The difference is in a world where all
10:56labor is being automated away, what is
10:59more valuable than the very industry
11:02that is automating it all away? If the
11:04danger to the economy is that AI is
11:06going to replace you, well, if you could
11:09pick one industry to be in, what
11:11industry would it be in? Would it be in
11:13accounting or would it be in,
11:15you know, the industry that does the
11:17replacing? Well, my money's on the
11:19industry that does the replacing. And so
11:21this isn't really a thesis.
11:23You know, it's it's a fact.
11:24Realistically, these are the systems
11:27that are automating away the work of
11:29everybody else.
11:30It's kind of why we're paid, right? The
11:31whole idea of automation is that it the
11:33use of technology to reduce labor. I've
11:35memorized that definition. Um
11:37So, in a world where virtually all
11:40arbitrage windows are shrinking, and in
11:42addition to that, virtually all
11:43knowledge work is being automated away,
11:45is there any industry you would rather
11:47be in than the one that is literally
11:48doing the automating? AI automation has
11:50probably the most staying power out of
11:51any industry on Earth. We are literally
11:54the irrigation canal by which the
11:56upstream river, you know, uh irrigates
12:00and then feeds the seeds that are the
12:02economy. We are the fingers of the hand
12:05of the brain of, you know, the wider
12:07economy. We are basically the things
12:09that are going out there and interacting
12:10with the businesses, uh you know,
12:12upskilling them, teaching them, building
12:14systems for them, and so on and so
12:15forth, that enable their automation.
12:18And I think if you deeply understand
12:19this, you'll also deeply understand what
12:21a responsibility we all have. Not just
12:23to do this responsibly, but to do this
12:25as effectively as humanly possible. I
12:27mean, we are tasked with basically being
12:29the flag bearers from economy 1.0 to
12:32economy 2.0. We are basically boots on
12:35the ground bringing about this new
12:37rebellion, this new this new way of
12:39thinking, way of life.
12:41You know, the future economy will not be
12:42mediated by people. It will be mediated
12:45by machines that are the successors and
12:47descendants of the sorts of systems that
12:49you and I are building for people right
12:50now. And so, if you think about it this
12:51way, if you think about it as like a
12:53like like like an existential no noble
12:56task, if you think about it as like
12:57something that is almost God-given to
12:59you, you know, I think you'll probably
13:00find that AI automation is not something
13:03worrying to be in or something scary.
13:05Like, it's actually probably the dopest
13:06ever, pardon my French. So, here's
13:08what I predict is going to occur over
13:10the next little bit. And here's what I
13:11would recommend you guys do as people
13:13that are obviously, you know, wanting to
13:15improve their ability to grow in an
13:16environment like this.
13:18The first is that I predict that over
13:20the course of maybe like the early half
13:22of 2027,
13:23probably not all of 2027 cuz things
13:25accelerate, but over the first half of
13:272027, a big portion of humanity is going
13:29to wake up to the fact that AI has
13:31totally blindsided them and is capable
13:33of way more than they ever gave AI
13:35credit for. The little chatbot can now
13:37build my entire business. What the hell?
13:40That sort of effect for us is not
13:42actually a bad thing because a big chunk
13:43of those people that are going to be
13:45having those realizations are going to
13:46be business owners.
13:48People that have lots of money but don't
13:50have a lot of time because their heads
13:52were down in their businesses this whole
13:53time. That's why they're only figuring
13:54out about it now.
13:55And so when they duck their heads up,
13:57take a breath of oxygen, they're going
13:59to look around and see how different the
14:00landscape has shifted and think, "Holy
14:02crap, I need to figure this stuff out."
14:04And who are they going to want to work
14:05with? Are they going to want to work
14:06with a machine? Or are these older
14:08school business owners, the ones that
14:09are busy generating capital with old
14:11school systems that they know, going to
14:12want to work with a real flesh and blood
14:14human? Obviously they're going to want
14:15to work with a real flesh and blood
14:16human, right?
14:17At least initially. And so
14:19there'll be a massive surge in interest
14:21for people like ourselves that can help
14:23them implement these systems. The caveat
14:25is what that help looks like is a little
14:27bit different than what it is up until
14:28now. What has been up until now is
14:30usually implementing like, you know, AI
14:32infrastructure. So some sort of like
14:34lead gen system or I don't know, ad
14:36creative generation system or, automated
14:38proposal generation system or, you know,
14:41pipeline delivery systems or, I don't
14:43know, ERPs or whatever the hell. Um in
14:45the future what it's probably going to
14:46look like is actually more like
14:47strategic um high-level advice. In a
14:49world where everybody can build
14:51virtually anything, it's not important
14:53how fast or how capable you are at
14:54building. What is is like, what are you
14:56going to build in the first place? Like
14:57how are you actually going to like build
14:59the, you know, irrigation canal of the
15:00future?
15:01And so in the first half of 2027 we'll
15:03probably see a big boost of this. Uh the
15:04reason why as well is because a bunch of
15:06people that aren't doing it are going to
15:07look around at their competitors that
15:08are doing it who are crushing them and
15:10they're going to think, "Holy I
15:11got to start juicing to keep up." And
15:12they're going to look for the nearest AI
15:13automation person to assist them with
15:15that. Um after 2027, I don't know. I
15:18wish I could say I knew. But with
15:20technology increasing so exponentially
15:22fast, I think it'd be intellectually
15:23dishonest for me to sit here and say,
15:25"Oh yeah, for sure X, Y, and Z will will
15:27will be a sure fire bet and you should
15:28absolutely keep doing X, Y, and Z."
15:30There are a lot of people on the
15:31internet that are going to pretend that
15:32they know. The unfortunate thing is in
15:34times of distress and also in times of
15:35great change, people usually flock to
15:37those that seem like they know, but the
15:39reality is most of them do not know. And
15:41so hopefully there is at least a little
15:42bit of intellectual humility there and
15:43you guys will give me the grace to just
15:45say, "I don't know. Industry changes
15:47pretty damn fast." Regardless, what
15:48industry would you rather be in than the
15:49one that is literally doing the
15:50automating, right? So whether or not
15:52there's a bigger boost in interest post
15:542027 or whether that interest dies off a
15:56little bit and starts to decline at the
15:57end of 2027, I think that's up for, you
16:00know, men in the sky with a big beard
16:02and maybe a lightning bolt to decide. Um
16:04what do you do about it now? Obviously.
16:07Why do I always bury these things at the
16:08end of the video? I don't know. Well,
16:09we're going to see a phasing out of old
16:12school drag and drop builders like
16:14make.com, Zapier, um and A&N, Power
16:17Automate, and so on, almost entirely. In
16:19Maker School, which is my um AI
16:21automation program where I basically get
16:22you up and running and get guarantee
16:24your first customer in 90 days. We have
16:25people that run like analyses on the
16:27market. They run constant analyses on
16:29freelance platforms. They run analyses
16:31of a variety of different things. But um
16:33we've seen consistent declines in the
16:36total project ratios of drag and drop
16:40builder tools relative to new markets
16:43like Claude Code, Codex, um you know,
16:45internal web apps,
16:47uh uh you know, systems, pipelines,
16:49consulting, and so on, etc. And so we're
16:51likely going to continue seeing that
16:53unless one of these wonderful platforms
16:54has like a total change of heart and
16:56actually become super awesome and
16:57fantastic, which I do not consider
16:59likely. The next thing that I would
17:00recommend is moving your outputs to a
17:03higher bandwidth medium. What I mean by
17:04this is previously it was enough to
17:07build a system that just generated text,
17:08let's say, or maybe some sort of simple
17:10imagery in a proposal or whatever. But
17:12because of the rising floor, today there
17:15is a a large market and there will
17:17continue to be an even larger market for
17:19um higher bandwidth modalities. So,
17:21think about text to image. You know,
17:22they say a picture is worth a thousand
17:23words. Well, you know, video is worth 24
17:25pictures a second. Therefore, it's worth
17:2724,000 words a second. Um I don't know.
17:29How about some sort of 4D Gaussian
17:31splat? That's probably worth even more.
17:32How about like some sort of
17:33three-dimensional visualization or some
17:35sort of, you know, simulation tool that
17:36allows somebody that installs air
17:38conditioning panels or whatever to like
17:39actually get in there. Stuff like this
17:41is going to grow increasingly valuable.
17:43And so, this is really like the place
17:44where you should be meeting people if
17:45you're selling implementation. Now, on
17:47the implementation bit, the industry is
17:49likely going to change from primarily AI
17:51implementation services like we're
17:52seeing right now, where you actually
17:54build the systems for the clients, where
17:55you connect the endpoints, you know,
17:57read the API docs, do all of the the
17:58funky stuff, and then, you know, give
18:00them a working system, to more education
18:04services, more uplifting services. So,
18:06think less AI implementation. Keep in
18:08mind, this is not all or nothing, one to
18:10zero. This is like, you know, 70% of all
18:12work to like 50% or 40% of all work. Uh
18:15but it's going to shift from AI
18:16implementation more to upskilling. It's
18:18going to shift to workshops. It's going
18:19to shift to team training. It's going to
18:21shift to um you know, essentially like
18:23consulting on companies and then being
18:26embedded within a company to help, you
18:27know, improve their teams as people gain
18:30more ability to do the implementations
18:31themselves and merely need somebody to
18:34help get their team up to speed on that
18:35fact as well. And so, we recently added
18:37a big like workshop uh portion to Maker
18:39School specific specifically for that
18:41reason. Um I was getting a lot of
18:42workshop gigs, a lot of other people in
18:43Maker Square getting a lot of workshop
18:45gigs. And we're kind of just like, "Hmm,
18:46I wonder why so many people want
18:47workshops all of a sudden?" And I think
18:49it's a rumbling of what is to come here.
18:51Um you know, Aaron Yeager walking across
18:53No, I'm just kidding. I love that show.
18:55All of this again means that businesses
18:57will want less stuff to build and more
19:00direction around what it is that they
19:01should build in the first place. Um they
19:03should also be able to, you know,
19:04meaningfully cooperate in a team and
19:05they just got like a one-click config
19:07that helps them just get up and running.
19:09So, think stuff like, you know, Claude
19:10Co-work instead of Claude Code
19:12necessarily. One big thing that I've
19:14been doing recently is going into teams
19:15and then teaching them how to use Claude
19:17in a general sense, then teaching them
19:18specific tactical ways to use Claude
19:20co-work, and then I am sort of like the
19:23supreme hegemon of the work workshop
19:25session will use Claude code to build
19:27the skills for their Claude co-work
19:29admin who will then distribute that to
19:31all the team members and then upskill.
19:32I'll also give them like weekly
19:34schedules and ways to like upgrade their
19:36skills that everybody in the team has
19:38approximately the same level of AI
19:40automation capability.
19:41You know, because obviously naturally
19:42disparities arise and stuff like that.
19:44So, building offerings like that with
19:46that structure are likely to be the big
19:48cash cow over the course of I want to
19:50say like early to mid 2027. In addition,
19:53AI automation for short-term growth
19:55hacking opportunities will likely be
19:57super valuable.
19:58When I say short-term growth hacking
19:59opportunities, if you think about it in
20:01a more dynamic market, the frequency of
20:04opportunities is going to increase.
20:05There's going to be more opportunities,
20:06okay? But, the total lifespan of every
20:09one individual opportunity will also go
20:10down. So, there's going to be a lot of
20:12like short-lived market opportunities
20:13where I don't know if you guys know
20:14anything about trading, but like you can
20:15buy and sell it sort of like ride the
20:17ride the dip over and over and over and
20:18over and over again because there's just
20:20going to be a lot of people entering the
20:21market with quick needs that are going
20:23to be quickly satisfied by, you know,
20:24autonomous beings.
20:26There's also going to be a lot of large
20:28opportunities there with asymmetric
20:29returns if you're capable of like
20:31spotting something early, jumping in on
20:32it, and so on and so forth. So, we're,
20:35you know, going to like the abstraction
20:37now a little bit which I know people
20:39don't really give a about, so I'll
20:40reel it back in to, you know, something
20:42practical. Like we've seen over the
20:44course of the last 5 years and like I've
20:45talked about every single year since I
20:46started, you know, making content on
20:47YouTube, hell, really just working in
20:49business and and consulting, the
20:51technical skill aspect of business has
20:54virtually evaporated today. If you spend
20:57any of your time improving your
20:58technical skills, you need to spend 100x
21:02the same units of time focusing on
21:03improving your sales, marketing, and
21:06delivery systems. That's what you should
21:08really be focused on. You should not be
21:09focused on, "Wow, what's the best
21:11harness to use to coral my agentic
21:14engineering workflow?" You should be
21:16focused on, "Okay, how can we sell as
21:18many people as possible in the next, you
21:19know, 2 months? Uh what is our funnel
21:21looking like? How what are we leaking
21:23in? Uh where are people falling off our
21:25funnel? What sorts of wording am I
21:26saying on the sales call? How am I
21:28pitching my social proof? Uh you know,
21:29am I demonstrating authority? Are we
21:31asking for the close? Are we following
21:33up on the close? What sort of
21:34dispositions do we have for inbound ad
21:36campaigns? And so on and so forth." So,
21:38everything to do with sales and
21:39marketing is like 100x as valuable as
21:41technical skills. Last year, it maybe
21:43was 10x as valuable. Now, it's 100x as
21:45valuable. In the next year, it's
21:46probably going to be 1,000x as valuable
21:47cuz everybody on planet Earth is going
21:49to have technical skills. Not actually
21:50technical skills, but they're going to
21:51have the ability to, you know, simulate
21:52technical skills with agents. If you
21:54really want to compete in 2028, 2029,
21:56and 2030, you got to be focused on
21:58actually articulating the benefits of
22:00services and understanding just the
22:01mechanics of business in general. All
22:03this to say, realistically, 2027 will
22:05probably be the most profitable year of
22:08AI automation to date. But, it may also
22:11be the beginning of its decline in the
22:13market. Whether that decline takes, you
22:15know, 5 years, 10 years, 15 years,
22:17whether, you know, AGI makes a freaking
22:20black hole in the center of the Earth
22:21and it swallows us all up tomorrow, who
22:23knows? Um but, with the industry's level
22:27of commoditization and capability of AI
22:29agents to produce outputs similar to
22:31humans increasing, you know, things are
22:32obviously going to change and I figured
22:34I would help you guys out on that
22:36change, on that evolution. So, I didn't
22:38make this video cuz I wanted to scare
22:39people. I made this video precisely for
22:40the opposite reason because I wanted to
22:42help people and I also wanted people to
22:43grow towards a more effective business
22:45model.
22:46Um you know, I know a lot of the
22:47comments are probably not like that and
22:48they'll probably be like, "Oh, Nick,
22:49what the hell are you talking about?"
22:50because they just don't want, you know,
22:51fear, uncertainty, and, you know, doubt.
22:54But, I'm not fudding for no reason,
22:55guys. I mean, we have Navier-Stokes
22:57mathematical millennium uh problems in
23:00math being solved by agents. We have
23:01agents designing cool works of art now.
23:03Like, let's be freaking real here,
23:05right? Like, agents are growing
23:06extremely powerful and it would be
23:07remiss of you not to at least think
23:08about what that means for the next
23:10couple of years.
23:11So, if you wanted to make money with
23:12this business model, insert sales pitch
23:14here. No, you guys should check out
23:15Maker School. Um it's my AI automation
23:17program where I guarantee you your first
23:18customer in 90 days or I give you your
23:19money back.
23:20Um we have over 2,000 people in there
23:23that do this thing on a daily basis.
23:25They are selling AI workshops, they're
23:27selling AI consulting, they're
23:28constantly up to date on the best ways
23:30to deliver AI implementation, they share
23:32notes around sales, marketing, on what's
23:33actually working in our industry. We
23:34also give around $20,000 in discounts,
23:37meaning you will immediately at least
23:3810x your investment the moment that you
23:39walk in. And of course it's guaranteed.
23:41What better offer could you possibly ask
23:43for? AI automation by definition is the
23:45use of technology to reduce human labor.
23:48I would rather be in an industry that
23:50does the reducing of human labor than be
23:53in an industry that gets automated away.
23:55I'm sure you guys feel the same way.
23:56Catch all y'all in the next video. See
23:58you.

About this transcript

This page contains the full transcript of Here's What I'd Learn Instead of AI Automation in 2027 by Nick Saraev, generated from the public captions YouTube serves with the video. The transcript has 5,552 words across 816 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 translate the transcript into 100+ languages.

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, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.