Emanuele Frontoni - Intelligenza Artificiale e creatività umana nell’epoca dell’Agentic AI — Transcript
Full transcript
- 0:05Good morning, everyone.
- 0:08My name is Venus, I am a student at the
- 0:10Morea Vivarelli Institute, and today we
- 0:12have the honor of hosting Professor
- 0:14Emanuele Frontoni, a full professor of
- 0:16computer science at the University of
- 0:18Macerata and co-director of BRAI, which
- 0:20stands for Vision, Robotics, and
- 0:22Artificial Intelligence.
- 0:26Professor Frontoni is a highly prominent
- 0:30figure in the scientific field.
- 0:34His research focuses primarily on
- 0:37artificial intelligence, the study of
- 0:40human behavior, and augmented reality.
- 0:44He is the author of over 300 scientific
- 0:47publications and actively collaborates
- 0:50with agencies and institutions on
- 0:52innovation and technology transfer
- 0:55projects.
- 0:58Since 2021, he has also been included in
- 1:00the prestigious list of the world's most
- 1:02cited scientists.
- 1:05Since 2022, he has also served as the
- 1:08scientific director of the Neolab at the
- 1:11Niguarda Hospital in Milan.
- 1:18The theme of the conference will be
- 1:21artificial intelligence and human
- 1:23creativity in the age of the agent, or
- 1:25systems that are not only capable of
- 1:27responding but of acting and
- 1:29understanding actively and autonomously
- 1:32.
- 1:35Through concrete examples, we will
- 1:37explore the opportunities offered by
- 1:38these technologies, as well as their
- 1:40limitations.
- 1:43The ethical implications, especially in
- 1:45professional and educational fields,
- 1:47will also be discussed.
- 1:50Without taking any more time from our
- 1:52guest, I hand the floor over to
- 1:53Professor Emanuele Frontoni.
- 2:02>> Thank you.
- 2:04Thank you for this invitation, and thank
- 2:06you to Venus, who studied everything
- 2:08perfectly.
- 2:09She was far too kind in her introduction
- 2:11.
- 2:15I was recalling with Sergio Mustica, who
- 2:17accompanied me here and even found me a
- 2:19secret parking spot because parking in
- 2:21Fabriano wasn't exactly easy, that I had
- 2:23been here some time ago, when I believe
- 2:24the very first editions of this festival
- 2:26were just beginning, right in this room
- 2:28.
- 2:31So, I am happy to return, also because
- 2:33in the meantime—and I’d like to greet my
- 2:36colleague Comodi over there from the
- 2:39Polytechnic, with whom I’ve been linked
- 2:41for many years since I was at the
- 2:44Polytechnic—you truly deserve
- 2:46congratulations, because this festival,
- 2:48along with the one in Foligno, is
- 2:51becoming an important institution for
- 2:53this area.
- 2:55And thank you for hosting so many
- 2:56organizations in these territories.
- 2:59Also, I've come to Fabriano often
- 3:00recently, Venus, because we’ve been
- 3:02recording, and I discovered the
- 3:04existence of a very amusing group of
- 3:06characters, your fellow townsfolk, who
- 3:08are the stars of "Autovelox," this...
- 3:15...podcast/videocast you record, which,
- 3:17by the way, hosted Greg a few days ago,
- 3:19of the famous duo Lillo and Greg, who
- 3:21was here in Fabriano with them; and so
- 3:23we had a lot of fun with him as well.
- 3:28And by the way, Fabriano is also the
- 3:30cradle of one of the most important
- 3:32projects that I will tell you about in
- 3:34conclusion, which is leading us, right
- 3:37alongside Sergio, to carry out many
- 3:39artificial intelligence initiatives
- 3:41being tested in the world of schools.
- 3:46And this project called Paideia, I'll
- 3:47tell you a bit about it towards the end
- 3:49because I'm curious to have a dialogue
- 3:50with Venus as well, I’ll be the one
- 3:52asking her questions later.
- 3:54No, I'm joking, of course, just to see
- 3:56if she agrees with us on certain points
- 3:58.
- 3:59Now I will do a short initial piece
- 4:01where I tell part of this story. Then
- 4:04obviously we will stop and debate, so
- 4:06I’ll do it quickly.
- 4:08So, first concept: this is me.
- 4:14Back in the era of Ghibli Style, the
- 4:17epic beginnings of AI image creation.
- 4:22And what do you think is strange about
- 4:24this photo?
- 4:25Ignoring the fact that I'm no longer
- 4:27blond but have white hair, and that I
- 4:29was fit while here I'm a bit more out of
- 4:31shape—all those variations—what is
- 4:33strange about that photo?
- 4:35What do you say, guys?
- 4:37I have a primary school nerd crowd here
- 4:38.
- 4:40I'm looking at you, I'm not looking
- 4:42around.
- 4:43What do you say? What's strange?
- 4:48This photo is illegal because it's a
- 4:50clear copyright violation of Studio
- 4:52Ghibli, which is the entity that
- 4:55produces these comics and these designs
- 4:57.
- 4:58It is a studio that works, that makes
- 5:00these drawings.
- 5:01To create this photo, I gave a
- 5:03generative artificial intelligence a
- 5:06photo of myself, and this is also an
- 5:08important step, because when you give
- 5:11files, data, or PDFs to a generative AI
- 5:14or an AI agent—which I will clarify for
- 5:16you shortly—you are, especially in the
- 5:19free versions, giving away data.
- 5:23So often we cannot do it.
- 5:25We cannot do it in the school world, we
- 5:27cannot do it in the corporate or
- 5:28professional world.
- 5:30On the other hand, I asked it to
- 5:32generate a photo in Ghibli Style; I
- 5:34could have asked it to generate a photo
- 5:36in a Spider-Man costume, it would have
- 5:38been the same thing.
- 5:42And this concept implies the fact that
- 5:45today one of the major concerns we must
- 5:47have regarding this world is also the
- 5:49privacy of our information and
- 5:51intellectual property.
- 5:55Which sometimes also means the privacy
- 5:57of your information and intellectual
- 5:59property.
- 6:00Now I will speak for a bit to the
- 6:02so-called adult generation present here
- 6:05—I see a few white hairs—as well as for
- 6:08a bit to the other, very young component
- 6:11.
- 6:11And we will try to play a game to
- 6:13measure the differences.
- 6:15And
- 6:17mind you, the lecture is not for them.
- 6:22Their generation has private Instagram
- 6:26accounts in 86%of cases.
- 6:30Our generation has public Instagram
- 6:33accounts in 92%of cases.
- 6:37Then their generation has 2.7 Instagram
- 6:38accounts, and we still haven't quite
- 6:40figured out what they do with them. How
- 6:42many Instagram accounts does Venus have
- 6:44?
- 6:46Venus has two.
- 6:47Someone, to reach the average of 2.7,
- 6:49must have three.
- 6:51The private, the very private, the
- 6:53private, okay?
- 6:55I won't make you raise your hands, but I
- 6:57think anyone who has three knows it too
- 6:59, okay?
- 7:02And we struggle to manage one.
- 7:07My mother, at one point, last Christmas
- 7:11, had an Instagram account with 189
- 7:16followers.
- 7:20My oldest daughter is 15, so she has
- 7:22access to Instagram, the second is 13
- 7:24and didn't have access to Instagram at
- 7:26the time, the third is seven.
- 7:29Guess whose Instagram account Grandma
- 7:31Gianna's was?
- 7:33Okay? Not exactly hers.
- 7:36The second daughter tried to defend
- 7:38herself, but obviously, retaliation was
- 7:39triggered against this fake Instagram
- 7:41account on Grandma's cell phone.
- 7:44There are grandparents among you here
- 7:46too, so be careful, eh?
- 7:47Okay? Because then we can move on.
- 7:51So, now you students stay still, I'm
- 7:55questioning the adults.
- 7:57Can you tell me how many of you know
- 7:59Photomath?
- 8:01Students, no, eh? Hands down.
- 8:04How many know Photomath?
- 8:09You almost do, though, eh, too young.
- 8:12Alright kids, now you answer, how many
- 8:14know Photomath?
- 8:21So I would start taking the lecture on
- 8:23studying for ourselves.
- 8:27Okay? Because either we start knowing
- 8:29this world a little, or we won't even
- 8:32know what to talk about, eh?
- 8:35We won't even know how to engage,
- 8:37because we are an educating community,
- 8:39eh? It’s not just the teacher who has to
- 8:41take care of them, eh?
- 8:44Please, we are all an educating
- 8:46community, everyone.
- 8:50Now what does Photomath do?
- 8:55If we talked to them or to Venus, they
- 8:57would tell me, "Teacher, it's a bit old
- 8:59now, this stuff is a bit out of fashion
- 9:01, it doesn't work very well, we have
- 9:03better things to use."
- 9:06But with this cell phone, I can take a
- 9:08photo of a math problem, a generative AI
- 9:11solves all the steps of the math problem
- 9:13, I copy them into my notebook for the
- 9:15teacher.
- 9:19Okay?
- 9:20Are there any math teachers by chance?
- 9:23I don't see any hands raised.
- 9:25But math teachers sometimes notice,
- 9:27sometimes they don't.
- 9:31But the strangest thing that happened to
- 9:34me regarding Photomath is that I was at
- 9:36dinner with a manager of a digital
- 9:38company; I tell this story often, and I
- 9:41said to her, she has three children like
- 9:43ours, the same age groups: 15, 13, 7.
- 9:47Only the genders are different. We have
- 9:49girl, girl, boy; she has boy, boy, girl
- 9:51.
- 9:52She is also our friend, but she is a
- 9:54very relevant figure in Italy, so I said
- 9:56to her, "Well, but yours must use
- 9:57Photomath too."
- 10:00We were at dinner and she replied to me
- 10:01, "I've never heard of it at home, so in
- 10:03my opinion, it doesn't exist here."
- 10:05A terrible sentence, isn't it?
- 10:08One that should really make so-called
- 10:11adults reflect quite a bit.
- 10:13We often think we are the holders of
- 10:16knowledge, yet we know almost nothing
- 10:18about that digital world where we all
- 10:20now live so extensively, and we know
- 10:22almost nothing about that world of
- 10:25artificial intelligence in which we all
- 10:27live quite extensively.
- 10:31To which I asked, I told her, "Look, in
- 10:33my opinion, I wouldn't trust this answer
- 10:35of yours too much, let's try writing to
- 10:37the family WhatsApp group," so she wrote
- 10:39to her children on the group and asked,
- 10:41"Do you know Photomath?"
- 10:44Okay?
- 10:46After 30 seconds, the middle child, a
- 10:47thirteen-year-old in eighth grade, wrote
- 10:49, "Yes, but I rarely use it."
- 10:52Okay?
- 10:54Which is the same answer they would give
- 10:56, okay?
- 10:58If questioned, they would say, "No, it
- 11:00doesn't work that well," or perhaps
- 11:02later they will ask me some questions
- 11:03and say, "Oh, I'm asking for a friend,"
- 11:05but I've had things of this nature
- 11:07happen to us in class.
- 11:10To this, I add that that generation—
- 11:12this generation here—has a very clear
- 11:15understanding of a new concept that we
- 11:17don't have clear in our minds, which is
- 11:20called human-AI teaming.
- 11:23That is, how do we have active
- 11:25collaboration between humans and
- 11:27artificial intelligence without
- 11:28delegating passively, but trying to
- 11:30build something interesting with it.
- 11:35I’ll add another piece of the story to
- 11:36this.
- 11:37I was at a conference like this one at a
- 11:39school
- 11:42and the teachers bring me onto the stage
- 11:45and tell me, "No, we are one of the best
- 11:47schools in Italy here, a model school;
- 11:50we've blocked the Wi-Fi, you can't
- 11:52access ChatGPT or Photomath, and we have
- 11:54a school-family registry, great
- 11:56communication with the family."
- 12:01I go up on stage and make a few jokes
- 12:04against prohibitionism, against the idea
- 12:06of "turn everything off, let's
- 12:08disconnect from everything, stop doing
- 12:10everything," because that's a rather
- 12:12strange idea, okay?
- 12:16But since we want to educate ourselves a
- 12:18bit within this journey, I suggested
- 12:20some paths, as I will do later with you
- 12:23, on the creative use of these tools.
- 12:27The students, who are extremely
- 12:28interesting and extremely intelligent,
- 12:30were waiting for me when I got off the
- 12:31stage and said, "Professor, can we talk
- 12:33to you?"
- 12:35And there were 30 of them waiting for me
- 12:37. They said, "Professor, don't believe
- 12:39what they say, okay?"
- 12:41"We have VPNs, we use ChatGPT from
- 12:43morning till night, Photomath, and not
- 12:44only that—the math teacher once left his
- 12:46computer open, we swiped his password,
- 12:48and now we give ourselves some of our
- 12:50own grades."
- 12:52But not only that, they told me: "Prof,
- 12:55we made a beautiful artificial
- 12:56intelligence agent that does Latin
- 12:59translations."
- 13:01And when you start using it—this is a
- 13:03good idea, guys, you can use it too—
- 13:05when you start using it, it asks what
- 13:07grade you got last time, because you can
- 13:09only improve by half a point at a time;
- 13:11otherwise, someone who always got a five
- 13:13and suddenly gets a nine stands out too
- 13:15much, and the teacher catches us.
- 13:19Okay? So they made this agent that makes
- 13:21everyone's translation a little better,
- 13:23but not too much.
- 13:24It adds a few mistakes in there—don't we
- 13:26want to reward them for that?
- 13:28Isn't this creativity?
- 13:30No?
- 13:31Would we have ever thought of this stuff
- 13:33?
- 13:35Absolutely not.
- 13:38And this is their great ability, which
- 13:41we must try to train a bit more,
- 13:43balancing awareness and appreciation;
- 13:45because if the issue is when they come
- 13:48to us adults or to us, the educational
- 13:51community, if Venus comes to me and says
- 13:54, "Prof, do you know Ambra, 8-bit,
- 13:56Angelo, Ash, and Bell?"
- 14:00Eh, eh, stop again, let's take the test
- 14:03. Adults, you guys stay still, eh.
- 14:07Who knows these names?
- 14:11No, no, guys stay still, hands down, the
- 14:13...
- 14:15None of the adults?
- 14:18Really, no one at all?
- 14:21Now, guys, do you know them?
- 14:25You're pretending, aren't you?
- 14:29How many trophies do you have?
- 14:33They are the Brawlers, the Brawlers from
- 14:36Brawl Stars. Okay? I have 3,000 trophies
- 14:39, eh?
- 14:41Oh well, I'm bad at it. What can I do?
- 14:43I'm bad at this.
- 14:45I have little time, let's say, to play
- 14:47with my kids. How many trophies do you
- 14:48have? Tell me.
- 14:5030,000, who is it?
- 14:53Did I hear a higher number?
- 14:5630,000?
- 15:00How many?
- 15:02>> 100.
- 15:04>> Who said 100?
- 15:06No, not 100, come on.
- 15:08100 is...100,000 is too many. Excuse me
- 15:10, I'm standing up, but I wanted to see
- 15:12you for a moment.
- 15:13Eh, so, this is the most played video
- 15:16game in the world in recent years.
- 15:21And by far, eh? Why is it interesting
- 15:23from the point of view of artificial
- 15:24intelligence? And I'm saying this to the
- 15:26teachers too, eh? When you go back to
- 15:28class, ask them how many trophies they
- 15:30have, and what they are doing with this
- 15:32artificial intelligence.
- 15:34Let's have a dialogue about this digital
- 15:36world, eh? I urge you, let's not label
- 15:39it as something different and unknown to
- 15:42us.
- 15:44My mother, the same one with the
- 15:45Instagram account, when I withdraw
- 15:47tomorrow afternoon after Sunday lunch to
- 15:49finish writing something for Sergio,
- 15:54she tells my children, "Your father is
- 15:55over there playing with the computer."
- 15:58Okay?
- 16:00My mother can easily afford this, but we
- 16:02, the parent class, the educational
- 16:04community, cannot afford it.
- 16:07We need to be a bit closer in our form
- 16:09of dialogue as well. Why does Brawl
- 16:11Stars have anything to do with AI?
- 16:14Because all those objects are agents,
- 16:17and they are agents that adapt to the
- 16:19game.
- 16:20Often they are humans, you know. You
- 16:22often play against each other as well.
- 16:26But when you aren't playing against each
- 16:28other, there are autonomous agents that
- 16:30don't beat you too brutally or win
- 16:31against you too quickly.
- 16:34Sure, does this cause problems?
- 16:36Absolutely.
- 16:39But while we talk about problems, we
- 16:40must, in my opinion, also build more
- 16:42awareness of what lies behind them.
- 16:44Look, there is a lot of interesting math
- 16:47inside Brawl Stars; there are a lot of
- 16:49interesting themes, but also interesting
- 16:52storytelling behind this game.
- 16:55Now in Italy, a lot of interesting
- 16:57artificial intelligence is being
- 16:59developed.
- 17:00The one you see behind me is Leonardo.
- 17:04I, Gabriele, and many colleagues use it
- 17:07. It is the largest supercomputer we
- 17:08have in Italy and one of the top 10
- 17:10supercomputers in the world.
- 17:13That's not nothing, you know.
- 17:15It's a two or three-hour drive from here
- 17:17, in Bologna, at Cineca, and in there we
- 17:20create many of the solutions that I will
- 17:22tell you about shortly, which are also
- 17:24the solutions that allow us not only to
- 17:27be passive users but to be active
- 17:29creators.
- 17:31Because look, guys, you have excellent
- 17:34ideas.
- 17:35Sometimes we don't understand them, so
- 17:37you have to insist a bit to make us
- 17:38understand, just as conversely, we try
- 17:40to explain concepts that you sometimes
- 17:42don't understand and that we insist on
- 17:44making you understand.
- 17:46But we are in the perfect phase of equal
- 17:48exchange between generations.
- 17:51That has never happened before, you know
- 17:52.
- 17:54You have a lot of things to teach your
- 17:56parents and even your teachers.
- 18:00And I urge you, the pact we make is that
- 18:02after today, you commit to talking a bit
- 18:04more about how this world of artificial
- 18:06intelligence works at school, at home,
- 18:08and around the world, okay?
- 18:11If you don't take charge of their future
- 18:12, there is no hope, you know.
- 18:15Don't hold it against me, you know, in
- 18:17this generational battle, okay? You are
- 18:20our allies.
- 18:23Because the idea of labeling the entire
- 18:25digital world as negative is something
- 18:27we oppose.
- 18:29We must understand it, and in all tools
- 18:30, there are positive things and negative
- 18:32things.
- 18:34And it depends on us, you know, on you—
- 18:36I don't know how you'll use it—on me, on
- 18:38them, and on how we build new ideas.
- 18:41That is why I insist that we must be
- 18:43builders of artificial intelligence, not
- 18:45just users of artificial intelligence.
- 18:48And mind you, we are building a lot of
- 18:50it.
- 18:51Now I will go on.
- 18:52We can do all this on one condition.
- 18:55That we stop humanizing it.
- 18:58Mistaking it for a human being, thinking
- 19:00it is more intelligent than a human
- 19:01being—that is not the case.
- 19:03And it has nothing to do with the human
- 19:05brain.
- 19:07This morning, before coming here, I was
- 19:08at a clinical conference on the use of
- 19:10music in the operating room.
- 19:13They were talking about the brain's
- 19:16reaction.
- 19:18And I was talking about artificial
- 19:19intelligence.
- 19:20At a certain point, a clinician said
- 19:22that we were probably talking about the
- 19:23same thing. No, we are absolutely not
- 19:25talking about the same thing, not even
- 19:27in our dreams.
- 19:28Our artificial neurons, those inside
- 19:30neural networks, are just little
- 19:31mathematical functions.
- 19:34They have nothing to do with a
- 19:35biological neuron.
- 19:37Absolutely nothing; it is just a
- 19:39metaphor.
- 19:40We used the same name, but it is not
- 19:42even true that we are inspired by the
- 19:44operating mechanisms of the human brain
- 19:46.
- 19:47Look, neural networks learn, for example
- 19:49, from massive amounts of data. The
- 19:51human brain does not learn like that.
- 19:55It is not like my son, my third one who
- 19:57is 7 years old, had to see 100,000
- 19:59photos of dogs, horses, monkeys, and
- 20:01cats to recognize them.
- 20:03A neural network must see 100,000 photos
- 20:05of dogs, horses, monkeys, and cats
- 20:07before it recognizes that animal.
- 20:12And by the way, the thing that surprises
- 20:13me a lot is that many times—and I say
- 20:15this especially to the adults, you guys
- 20:16a little less—according to the
- 20:18statistics we have, you thank it.
- 20:20You tell ChatGPT: thank you, good
- 20:23morning, great job.
- 20:25Ah, the deal we make is that from today
- 20:27, we stop.
- 20:30We don't thank it anymore, we don't
- 20:33address it as if it were a human being,
- 20:36and you, the younger generation, help us
- 20:39spread this and stop discussing personal
- 20:42matters with artificial intelligence
- 20:45systems.
- 20:51I repeat the phrase: stop discussing
- 20:53personal matters with artificial
- 20:54intelligence systems.
- 20:57Look, adults are there, they listen to
- 20:59you, sometimes they are distracted, but
- 21:01that system is a statistical system.
- 21:04If you ask it how to approach that guy
- 21:06or that girl you like, it answers
- 21:08differently every day, and the risk is
- 21:10that the other person will think you're
- 21:12crazy if you change your mind and your
- 21:15strategy every day.
- 21:17Okay? I'm counting on you.
- 21:19I understand it listens to you, it's
- 21:21always available, it doesn't tell you to
- 21:24wait a minute, it doesn't judge you, it
- 21:26doesn't tell you that what you're doing
- 21:28is wrong, but it has no knowledge of
- 21:31your psychological sphere.
- 21:34Okay?
- 21:35The data we have is very worrying; 68%
- 21:38of adolescents use AI systems to discuss
- 21:41things within their own psychological
- 21:44sphere.
- 21:46That stuff risks answering you with
- 21:49absolute nonsense.
- 21:52Okay?
- 21:54We cannot take that risk, so yes, use it
- 21:56, discuss things, do whatever you like
- 21:58with it, but remember that meaningful
- 22:00dialogue, especially in the
- 22:02psychological sphere, happens between
- 22:04human beings because that object knows
- 22:06nothing about the human brain; it has
- 22:08read something, but it doesn't know the
- 22:11entire context surrounding you.
- 22:16By the way, and we are in Fabriano, so
- 22:18this example is quite fitting, to remind
- 22:21us that we have never engineered
- 22:23anything—we engineers, me, Gabriele, and
- 22:25many others here—like a human being does
- 22:28, to remind us that it is not human
- 22:30intelligence, it doesn't resemble it at
- 22:33all.
- 22:35Think about the washing machine.
- 22:37When the washing machine was made, even
- 22:40here in Fabriano, they didn't say, "
- 22:42Look, there's a person there turning a
- 22:44drum 100 times a minute to wash clothes
- 22:46, let's copy him or her," but we
- 22:47engineered it in a different way.
- 22:51We invented another way to do it. Well,
- 22:53artificial intelligence is just another
- 22:56way to analyze data and it should not be
- 22:58humanized, mind you.
- 23:00Because we adults make a very big
- 23:02mistake there, too.
- 23:04Actually, think about it: we write "
- 23:06thank you" to it.
- 23:08After it has written all its things, we
- 23:10write "thank you."
- 23:12We activate a model with 100 billion
- 23:14parameters.
- 23:15Hmm? A remote server between Ireland or
- 23:18the United States.
- 23:21We consume a lot of electricity, we heat
- 23:22up a lot of fresh water just to get "
- 23:24you're welcome" as a response.
- 23:27Which doesn't seem like a great
- 23:28acquisition of knowledge to me. Eh?
- 23:33So, I won't ask for a show of hands of
- 23:35who thanks ChatGPT, but the other pact
- 23:37we make today is that we stop using it
- 23:39as if it were human.
- 23:43Look, those of us who have programmed
- 23:45our whole lives in Python, in C, in C ++
- 23:48, when we wrote Python code, we didn't
- 23:50say, "Excuse me if I'm bothering you,
- 23:52I'll stop bothering you now, I won't
- 23:54take up too much of your memory," we
- 23:56were very, very direct.
- 24:00Okay? So, please, let's stop thanking it
- 24:03.
- 24:04Because that is a machine and we are
- 24:05human beings, and we must understand
- 24:07what that machine does and what we do as
- 24:08human beings.
- 24:10By the way, I have never seen you thank
- 24:12your toaster or your coffee machine
- 24:14after it finishes a job, so it escapes
- 24:16me why you thank a machine.
- 24:19Look at this.
- 24:20This is Moltbook.
- 24:22Have any of you taken a look at it?
- 24:26It's a very recent social network, born
- 24:29a few weeks ago, where many artificial
- 24:32agents can chat with each other and we
- 24:34humans are just watching from the
- 24:37outside.
- 24:39See, it's written down there.
- 24:45Put that way,
- 24:47it caused an uproar, okay? Because
- 24:51imagine a social network where the only
- 24:53ones who can exchange likes, chat, and
- 24:55comment are artificial agents.
- 24:58The journalists jumped on it immediately
- 24:59.
- 25:01I’ve done at least 10 interviews a week
- 25:03over the last few weeks about Moltbook,
- 25:05but in reality, there’s nothing to worry
- 25:08about.
- 25:10It's a nice exercise; we even made two
- 25:13agents that talk on Moltbook, one chats
- 25:15about soccer and the other about
- 25:17referendums, which was a rather crucial
- 25:19topic until a few weeks ago.
- 25:22All agents that chat with each other
- 25:24degenerate within a few hours.
- 25:27What does degenerate mean? They all talk
- 25:29about the same subject, they all state
- 25:31the same theory, and they all have the
- 25:32same idea.
- 25:33But it's normal; think of ChatGPT when
- 25:35you tell it, "Look, you made a mistake,
- 25:37" and it replies, "Oh, of course I made
- 25:39a mistake, tell me how I can correct
- 25:40myself."
- 25:42So now everyone says Inter will win the
- 25:44league—which doesn't take much to say—
- 25:47and that "no" will win the referendum,
- 25:50and we know that by now too.
- 25:55I repeat, artificial agents that chat
- 25:57with each other degenerate in a few
- 25:58hours; they all say the same thing
- 26:00because those tools aren't creative,
- 26:02it's not like someone at a certain point
- 26:04says, "Oh, I have a different idea."
- 26:07"Go ahead, I'll say it, it's different
- 26:09from the others." This concept does not
- 26:11exist. We humans are the creative ones,
- 26:13eh.
- 26:15And don't give me remarks like the
- 26:17journalists who always tell me, "Oh, but
- 26:19they are inventing a secret language and
- 26:21they will turn against us."
- 26:24No, that's Terminator, it's science
- 26:26fiction, it's wonderful, I'm a fan of it
- 26:27too, but it has nothing to do with
- 26:29reality.
- 26:30It’s not possible, they cannot invent a
- 26:32secret language.
- 26:34They don't have that creativity. If we
- 26:36train an artificial intelligence system
- 26:39to
- 26:41design bags. Venus, do you like bags?
- 26:45If we teach it to design bags and the
- 26:47bags we teach it, that we show it, are
- 26:49all rectangular bags, that system will
- 26:51generate rectangular bags for its entire
- 26:53life.
- 26:55It’s not like it wakes up in the morning
- 26:57and says, "Now I'll make a sack or a
- 26:59circular one," eh.
- 27:01Please, this must remind us that we are
- 27:03the creative ones and those are machines
- 27:05.
- 27:07At this point, that’s where the game is
- 27:08, eh. What do we let those machines do
- 27:10and what do we do?
- 27:12Well, look, to show that we can't do
- 27:14everything, look at this match.
- 27:18We swiped the robots from Gabriele’s
- 27:20department.
- 27:21No, I'm joking, but we tried to have a
- 27:24human play against a little robot at
- 27:27tic-tac-toe.
- 27:28What is it called, Fabriano?
- 27:30Eh?
- 27:30>> Tic-tac-toe.
- 27:31>> Tic-tac-toe.
- 27:32In the Fermo area it’s also called "
- 27:34fila fila tre," which is a fun name.
- 27:37To train it, we showed it only games
- 27:39with X's and dots, X's and dots, nothing
- 27:42else.
- 27:44Let’s see who wins.
- 27:47The human puts an X, the robot puts a
- 27:50dot, the human an X, dot, X, dot.
- 27:55It won.
- 27:57Okay?
- 27:59And it won correctly because it doesn't
- 28:02know the concept of a line or a border,
- 28:04since I told you we only let it see X's
- 28:06and dots.
- 28:08When that camera looks down and sees
- 28:10those horizontal and vertical lines, not
- 28:13having this information in its knowledge
- 28:15base, it doesn't deduce them or make
- 28:17complex reasoning because for us, we say
- 28:20it's obvious that it's the playing field
- 28:22and that the X's and dots go in there.
- 28:26For AI, the obvious doesn't exist, there
- 28:29is no so-called common sense reasoning.
- 28:33Oh, that’s no small thing, eh?
- 28:36Think about in a clinic, this morning we
- 28:38were with clinicians, one of the big
- 28:39problems in applying artificial
- 28:40intelligence in the clinic is that
- 28:42protocols don't say everything.
- 28:45Clinical protocols are well written for
- 28:47a competent human, not for a stupid
- 28:49machine.
- 28:51It doesn't say "while operating, don't
- 28:52cut the aorta because they'll die,"
- 28:54because it's obvious to a clinician that
- 28:56you shouldn't cut the aorta because
- 28:57they'll die.
- 28:58Okay? It's not equally obvious to a
- 29:00machine.
- 29:02So now we are working on making "
- 29:04machine-readable" protocols, we call
- 29:06them, where even the obvious is written
- 29:08down, because if the obvious isn't
- 29:09written, that machine doesn't know it.
- 29:13And look, that is no small limitation
- 29:15either, eh.
- 29:16But surely we can do interesting things
- 29:18. This is one of the examples I would
- 29:20have shown you.
- 29:21Let’s try, I tried while we were there
- 29:22quickly to play a video again, let’s see
- 29:24if it plays.
- 29:25You made it.
- 29:28A class made it, it's a specialized
- 29:29agent that speaks your language, you can
- 29:31tell it "hey bro, what's up?"
- 29:34Okay? And it answers you, it says "ah,
- 29:37I'm bubi bli," okay? And I'm here to
- 29:39keep you company. "Today I'm feeling
- 29:41like a 'maranza'," I'll write to it.
- 29:43And he replies again, "Oh, that's how I
- 29:45like it, bro," going out to the square
- 29:47with speakers blasting techno in your
- 29:49eardrums.
- 29:50Do you guys talk like that too?
- 29:51>> No.
- 29:52>> Huh?
- 29:53Total negative vibes.
- 29:57So how did this specialized agent come
- 29:59about?
- 30:01A good Italian teacher walked into class
- 30:02and told them, "First off, write me a
- 30:04glossary."
- 30:06Now my daughters tell me, "Great aura."
- 30:09Okay?
- 30:10Do you use "aura" in Fabriano, here in
- 30:11Fabriano?
- 30:12Huh?
- 30:13>> Yes.
- 30:14>> That's how they told me it's spelled,
- 30:15yeah, literally "aura."
- 30:17Huh? It's a thing, go ahead. What do you
- 30:19say, Venus?
- 30:19>> It is used.
- 30:20>> It's used?
- 30:20>> Yes, yes.
- 30:21>> What does it mean?
- 30:22>> Well, aura is like...
- 30:24>> Come closer, come closer so I can hear
- 30:25you better.
- 30:26>> Uh, when you do something incredible or
- 30:29do it well, they say "aura."
- 30:31>> So "great aura" is positive. Good.
- 30:34I'm happy.
- 30:35>> But also when you do something...
- 30:36>> bad, you lose aura.
- 30:37>> You lose aura. Come on, they tell me
- 30:39that too.
- 30:40Eh!
- 30:42So, the teacher entered the class and
- 30:45had them write a glossary; then they
- 30:47went data hunting, because the knowledge
- 30:50of these agents is created through data
- 30:53: for bags, bags for math exercises,
- 30:56math exercises for writing in this
- 30:58strange language.
- 31:01So many messages written by them in this
- 31:03language; where did they get them? From
- 31:05the class WhatsApp group.
- 31:07Two years of class WhatsApp, 23,000
- 31:09messages; we took them, anonymized them
- 31:11, and then fed them to this system.
- 31:16Oh, that's not a small feat, eh. This
- 31:17system, which is called Bubbl.bl, won
- 31:20all the awards for innovative Italian
- 31:22education we could possibly win last
- 31:24year. We won at Didacta, at the States
- 31:26General, at the Valdobbiadene Festival
- 31:29of Educational Innovation—everyone.
- 31:31This stuff was done by a class in
- 31:34Recanati at the Mattei, which is one of
- 31:36the institutes Giorgio works with, one
- 31:38of the people who works with us on
- 31:41Paideia.
- 31:43Look, this story is very creative and
- 31:45tells us a lot about how we will use
- 31:47agents. We won't be using generalist
- 31:50tools, but rather highly specialized
- 31:52ones.
- 31:53And this also applies to Industrial AI,
- 31:56eh. We are in Fabriano, one of the
- 31:58cradles of so much Italian industry, and
- 32:01much of this Industrial AI is born here
- 32:03with us helping them develop many of
- 32:06these specialized agents.
- 32:09Even robots—just yesterday, or the
- 32:11afternoon before, we were in a national
- 32:13webinar with Campetella showcasing our
- 32:15specialized agents for robotics.
- 32:19And this also holds true when we move,
- 32:21for example, to the human brain.
- 32:23Look, this is a recent study we
- 32:25conducted with Harvard; it almost, how
- 32:27should I put it, it always gives me a
- 32:29certain thrill to see our lab alongside
- 32:32this, which is one of the greatest and
- 32:34most important universities in the world
- 32:36.
- 32:38Emanuele, one of our team, ended up over
- 32:41there, and we started trying to
- 32:42understand if we are capable of
- 32:44reconstructing visual signals, acoustic
- 32:47signals, and multimedia signals starting
- 32:50from EEGs, from certain brain activation
- 32:52signals.
- 32:56And the amazing thing emerging from this
- 32:58beautiful research is that if I can
- 33:01manage to collect enough data (which is
- 33:04what AI needs), then I can trigger a
- 33:06generative artificial intelligence
- 33:09system and, within it, reconstruct that
- 33:11brain signal to see the two-dimensional
- 33:14and three-dimensional object that was
- 33:17created.
- 33:20Look, this is significant; it could mean
- 33:22solving how to bring sight to those who
- 33:24cannot see, bringing three-dimensional
- 33:27modeling capabilities to those with only
- 33:29one eye, or bringing hearing to those
- 33:31who cannot listen.
- 33:33Okay?
- 33:35Because it means knowing how to
- 33:36reactivate small areas of the brain that
- 33:37are the ones that generate [signals].
- 33:41So, is this interesting?
- 33:43It is very interesting.
- 33:45Could we do this before artificial
- 33:47intelligence?
- 33:48No, it was a very complex problem to
- 33:51solve.
- 33:54This artificial intelligence, used
- 33:56creatively and not as a passive
- 33:57delegation, is amplifying our research
- 34:00capacity—not because we write text
- 34:02faster, that’s not it—but because it
- 34:04allows us to solve problems that were
- 34:06previously computationally unsolvable.
- 34:11And this also applies to other
- 34:13applications.
- 34:14In this case, we are in the Macerata
- 34:17valley, and this is Giacomo, my third
- 34:20son, whom we asked to simulate
- 34:22stereotypies.
- 34:25Today, an autism therapist watches that
- 34:27child and measures how many repetitive
- 34:29movements they make.
- 34:32Then Giacomo is the third child, so he
- 34:34tries a bit, you see, he moves, makes a
- 34:36few motions, then leaves us there and
- 34:38walks away.
- 34:39Giacomo has no problems, eh; he was just
- 34:41used as a guinea pig for these actions.
- 34:43But it is very interesting to be able to
- 34:45measure it, isn't it?
- 34:46And you see that when we measure him,
- 34:47there are all those little dots up there
- 34:49.
- 34:49Why are there those dots?
- 34:51Because we want to explain to the
- 34:53clinician that the system is working.
- 34:56It’s called
- 34:59Today it is written
- 35:02because if I don't explain to you how I
- 35:04am working, how can you trust me?
- 35:08We don't want closed boxes that keep the
- 35:10human being out of this game, you know?
- 35:13But we want the human being to be a
- 35:14co-protagonist in this process.
- 35:17And look, still regarding autism, this
- 35:20is Eleonora, the daughter I mentioned
- 35:23before, the
- 35:26And you see she has dots drawn on her
- 35:28face because, once again, autism
- 35:30therapists told us that eye contact in
- 35:32autism is a very important issue.
- 35:35So you can say, "Eleonora, will you look
- 35:36at me?"
- 35:38If Eleonora looks at us, those little
- 35:40green lines align, we measure how many
- 35:42seconds they align, and so we know if
- 35:44she is improving in terms of eye contact
- 35:46.
- 35:47Fewer repetitive behaviors, more eye
- 35:49contact—the therapy is going well.
- 35:51Conversely, we must change our approach
- 35:52.
- 35:54Even here, you can see it works, right?
- 35:56And you can understand how it works; why
- 35:58are there blue dots and green lines?
- 36:01Because the human being on the other
- 36:02side needs to understand that the object
- 36:04is helping them take a measurement.
- 36:07The number 27 isn't written there for no
- 36:11reason.
- 36:14This, guys, also requires new forms of
- 36:16interaction, and these ideas don't come
- 36:19from me, but from the young people in
- 36:21our lab.
- 36:23They don't come from your teachers, but
- 36:24from you. Look, none of us would have
- 36:27thought of people improving their Latin
- 36:29translations by half a point at a time
- 36:31so as not to draw the teacher's
- 36:32attention.
- 36:34Okay?
- 36:36You are creative, you have a ton of
- 36:38skills, including in multimedia.
- 36:41Okay?
- 36:43When you see teachers arriving, or
- 36:44adults or older people, showing up with
- 36:46slides using Times New Roman text with
- 36:49letters dropping from above, run for the
- 36:51hills—the Flintstones have arrived.
- 36:56Then, of course, listen, study
- 36:58everything you can study—studying is
- 37:00very interesting—but add your own touch
- 37:02, have your say, and demand that these
- 37:04topics be discussed.
- 37:08I’m counting on you; it’s the pact we’re
- 37:09making this morning, okay?
- 37:14Now, who is helping us do all this?
- 37:19GTA V.
- 37:23Okay, again, all the adults don't even
- 37:26know what that is.
- 37:28Some do, the more nerdy ones; it's a
- 37:30video game.
- 37:31See, the audio and video technician
- 37:34nodded because he remembers it, right?
- 37:38Now the 6th one is coming, right?
- 37:40Is it coming?
- 37:42No?
- 37:43Some say it's a legend; I'm not really
- 37:46sure if it's true or not that GTA VI is
- 37:48coming.
- 37:50Why do we need a video game? We need it
- 37:53because all those little characters
- 37:55moving around are our datasets that we
- 37:57operate with.
- 37:59Okay?
- 38:01Now to conclude,
- 38:05we also need you for the experiments.
- 38:07This is Paideia.
- 38:10And Paideia is a very interesting
- 38:12experience because it brings together
- 38:14Laboratorio delle Idee, the Fabriano
- 38:16community, and schools—many schools—and
- 38:18in there, we are experimenting with how
- 38:20we will use specialized AI agents in
- 38:22school.
- 38:25We won't show it to you now because we
- 38:27only have slides, but if you tell that
- 38:30specialized agent to write you an essay
- 38:32on the character of Renzo in "The
- 38:34Betrothed," it will reply: "No, I won't
- 38:36write the essay, but if you want, I'll
- 38:39help you with some arguments or an
- 38:41initial outline of what you can write."
- 38:45And then it has many specialized tools
- 38:47because you have them; especially in
- 38:49your age group, it is a tool that can be
- 38:52interesting. It is also a tool for
- 38:53inclusion. Remember that the whole world
- 38:55is autism, and AI for autism is born in
- 38:57the Marche region.
- 38:59It is born in the Marche for the world,
- 39:01right?
- 39:02And so we would like a lot of AI
- 39:04specialized for AI to be born in the
- 39:06Marche. This is the Paideia manifesto.
- 39:09And Sergio, by the way, was its main
- 39:12actor, eh.
- 39:14Look at what is written. I will only
- 39:15highlight a few points. Effort and
- 39:18struggle have formative value.
- 39:23Look, you need to feel a little bit of
- 39:25struggle.
- 39:27I mean, everything that doesn't make us
- 39:29struggle is usually not very interesting
- 39:30.
- 39:32I am a big fan of cycling and riding in
- 39:34the mountains by bike.
- 39:36For weight reasons, as you can see, I
- 39:38use an electric bike, but it is much
- 39:40less fascinating. I would be much more
- 39:42satisfied to be able to use a real bike
- 39:44. You pedal harder, but the satisfaction
- 39:46is a thousand times higher.
- 39:49Down there it says "awareness," the true
- 39:52dividing line for the use of AI.
- 39:56It is what we talked about at the
- 39:57beginning of this morning. We are here
- 39:59because we want to generate awareness.
- 40:02If we are not aware, we are slaves to
- 40:04something else or someone else.
- 40:08AI at school has clear boundaries.
- 40:12That is, let's decide together how we
- 40:14will use it, talk about it, and then
- 40:16once we have established boundaries, we
- 40:18write them down and respect this
- 40:19educational pact.
- 40:21It cannot be the Wild West, it cannot be
- 40:24that everyone wakes up in the morning
- 40:26and does whatever the heck they want,
- 40:28because AI is by now an integral part of
- 40:30a piece of our digital life that is a
- 40:32piece of our real life.
- 40:35Let's not make these distinctions, and I
- 40:36say this to adults and the educational
- 40:37community.
- 40:39And in this game let's also remember
- 40:42that
- 40:45learning, point 3, read the title, eh,
- 40:47then you can find this online on the
- 40:48Paideia website, download it, bring it
- 40:50to class, do whatever you want.
- 40:53Learning means recognizing needs and
- 40:55limits.
- 40:56You are very different from one another
- 40:59.
- 40:59I am a barbarian of teaching; I walk
- 41:01into a classroom and I, Gabriele, ask at
- 41:04the university, "We have 150 students,
- 41:06let's tell them everything we know and
- 41:08then we leave."
- 41:10Your teachers are with you every hour,
- 41:13every day, every week for years with 25
- 41:16people, with all your diversity, and one
- 41:19of the great uses of AI in teaching is
- 41:22personalization.
- 41:26It’s being able to create material that
- 41:28is best suited for you to learn, for
- 41:30both the very capable and those less
- 41:32capable.
- 41:34And the whole world of pedagogy is
- 41:36pushing strongly in that direction.
- 41:39Which, however, does not mean shortcuts
- 41:41, but in my opinion, it's obvious that
- 41:43you look for shortcuts, we all looked
- 41:44for them.
- 41:47Now there is a very good Spanish teacher
- 41:49of my daughter's who forces all these
- 41:52kids to meet up in the afternoon, which
- 41:54is interesting, because when you are
- 41:56together you have fun, right?
- 41:59>> Sure.
- 41:59>> Sure, says Venus.
- 42:01To meet up together and record videos.
- 42:05Their only homework is the production of
- 42:08Spanish news programs, Spanish tourist
- 42:10tours, Spanish Manzoni, all videos
- 42:12recorded and edited by them.
- 42:16Look, you are world-class on this topic
- 42:18, you are world-class, you must be
- 42:20convinced of it, provided we study a
- 42:22little and agree to talk about these
- 42:24things together.
- 42:27Why do I insist on this, and I'll close
- 42:28with this?
- 42:29Because we need ideas, and look at this
- 42:32idea.
- 42:34This idea is one of the best we have
- 42:36carried forward in the last 2 years, and
- 42:39it is an idea linked to NEMOLAB, which
- 42:41is a laboratory that tries to respond—
- 42:43and we must keep the word laboratory in
- 42:46mind at school, eh. Paideia says, dear
- 42:49schools, you are a laboratory for
- 42:51experimentation.
- 42:52Let's discover together what we must do
- 42:54.
- 42:57And this, this story stems from the fact
- 42:59that we asked ourselves, what
- 43:01interesting things can we do in the
- 43:03world of artificial intelligence when
- 43:05dealing with an ALS patient?
- 43:08That is, someone who loses the use of
- 43:10vital functions like their voice?
- 43:13Let's see it. Is this with audio?
- 43:17I think the audio has started.
- 43:22If the control room can help us.
- 43:30So I repeat, the Nemo center also exists
- 43:32in Ancona, eh?
- 43:34For a few years now we have managed to
- 43:36open it here at Torrette too, and in the
- 43:38meantime, we try to solve and
- 43:43The Nemo center in Ancona deals with ALS
- 43:46and SMA.
- 43:47SMA is a sort of ALS for children.
- 43:51The birth of these Nemo centers happened
- 43:53at Niguarda, that center that comes from
- 43:54Niguarda deals with this.
- 43:57Now, while the video plays—the video is
- 43:58very interesting and also very well made
- 44:00—
- 44:01I will tell you about it in my own words
- 44:04.
- 44:06Today you can donate your voice to us,
- 44:08and you can donate your voice to allow
- 44:10us to collect data.
- 44:13This voice you donate to us is what
- 44:15allows us to recreate generative systems
- 44:18, AI agents that recreate the voice of
- 44:20that patient.
- 44:22It doesn’t take much, just a prompt.
- 44:24This prompt is usually an old voice
- 44:28message or an old video recorded on some
- 44:31old, uh, some old medium.
- 44:35From that audio, we can reconstruct that
- 44:37patient's voice. That voice, that
- 44:39algorithm, is inserted into the voice
- 44:41synthesizer.
- 44:42These patients communicate with you,
- 44:44Venus, through eye-tracking, and if you
- 44:46talk to them, it's like they're chatting
- 44:48just as we are chatting among ourselves
- 44:50, and they can finally use their own
- 44:52voice.
- 44:54When we presented the first results,
- 44:56some of these patients told us, "Prof,
- 44:58now I call my friends and my friends
- 45:00recognize me."
- 45:04Look, that same technology is what we
- 45:06use to fake the voice of an elderly
- 45:07gentleman's son to ask that gentleman to
- 45:09send us a wire transfer, to bring us
- 45:11money to his doorstep because his son
- 45:13had an accident, and we commit fraud.
- 45:17And this, unfortunately, is the story of
- 45:19technology. We can have negative uses
- 45:21and extremely positive uses.
- 45:24And it is up to us, uh, to create these
- 45:25ideas. You see, go when you are at
- 45:27school, or whoever wants to, to that
- 45:29site and donate your voice to us.
- 45:33This donation serves to collect data
- 45:35because AI feeds on data.
- 45:38Last year we collected 15,000 Italian
- 45:40human voices; we need all of them with
- 45:42the accents of Fabriano, of Matelica,
- 45:44from Fermo, like my Milanese one, and
- 45:46with this we have given back a voice to
- 45:48the first 96 Italian patients.
- 45:51That’s not bad, right?
- 45:52And this is a great idea, it’s a great
- 45:55idea that came from your generation.
- 45:58And we really need these kinds of ideas
- 46:00.
- 46:01So, to conclude, so I can hand the floor
- 46:04back to Venus, we have two possible
- 46:06reactions to this evolution—which I
- 46:08wouldn't call a revolution, it’s one of
- 46:11the many evolutions we face—:
- 46:16either stay there frozen, a bit scared,
- 46:18watching what happens, or get moving,
- 46:20use all the human creativity that
- 46:22distinguishes us and define a positive
- 46:24use of this artificial intelligence that
- 46:26isn't passive delegation, but rather an
- 46:28amplification of human capabilities, and
- 46:30then use it to improve people's quality
- 46:33of life and improve the quality of our
- 46:35society.
- 46:37This is a bit of our mission; I hope,
- 46:38Venus, that you will help us with this,
- 46:40okay?
- 46:41Thank you, good luck.
- 46:54>> I had a question.
- 46:57I mean, we have moved from an artificial
- 46:59intelligence that analyzes data to one
- 47:01that creates content.
- 47:03So what could be the skills that young
- 47:05people should acquire to, precisely, not
- 47:07be replaced by these technologies?
- 47:10>> Well, that’s a good question as you head
- 47:12back to class or home. I urge you to
- 47:15study in the broadest way possible.
- 47:19I mean, don't specialize, wait to
- 47:21specialize for as long as possible,
- 47:23study everything you can study because
- 47:25this is the only antidote we have to
- 47:27these possibilities.
- 47:30And this maintaining a broad culture of
- 47:33ours is something I also tell adults,
- 47:35okay?
- 47:37Because we too have stopped studying a
- 47:39bit and we can't afford that.
- 47:43>> Thank you.
- 47:46>> Thank you. Thanks Venus.
- 47:49Give her a round of applause, she was
- 47:50better at presenting.
- 47:57>> Thank you.
About this transcript
This page contains the full transcript of Emanuele Frontoni - Intelligenza Artificiale e creatività umana nell’epoca dell’Agentic AI by Festa Scienza Filosofia, generated from the public captions YouTube serves with the video. The transcript has 7,302 words across 1,249 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.