MIT: Imagination In Action’s-AI Summit Panels 2024 by Forbes - Dailymotion

Forbes

Watch the playlist MIT: Imagination In Action’s-AI Summit Panels 2024 by Forbes on Dailymotion

Transcript

00:00We are so excited to welcome you here for our symposium on the future of AI.

00:09And I just, I've known you for a number of years now.

00:13And I've known you to be as passionate about AI as you are about music.

00:18And so I wonder if you can share with us a little bit about your journey in AI and tech.

00:25How did you get interested and how did you come along for, alongside us?

00:33I got interested in this space here, in this space.

00:41In 2005, I came to Boston, Black Eyed Peas were on tour.

00:48And I put in on my contact, on my address book and on my phone.

00:55I always used to, and I still to this day, put in what people do.

00:58Like, hey, what do you do?

00:59Oh, I'm an engineer or I write songs or I'm a composer, I'm a choreographer.

01:06And I put the city that they're from.

01:08It's always just been like a habit, a practice.

01:13So there was a guy by the name of Alan Heardstone, engineer from Boston.

01:19So in 2005, I'm like bored here in Boston right before a show we had two days off.

01:25I'm like, who do I know in Boston?

01:28So I typed in Boston and out pops up Alan Heardstone.

01:31I was like, oh, that's right.

01:33That's that guy that I met at the studio, that engineer, nerdy dude.

01:41Let me see if we could hang out.

01:44Maybe he knows something to do in Boston.

01:47So I'm like, yo, what's up, Alan Heardstone?

01:48I met you back in, I think, 2000.

01:50He was like, hey, what's up, Will.

01:52I am.

01:53I never thought you were going to call me.

01:57I was like, well, you told me to call you if I'm ever in Boston, so here I am.

02:00He's like, oh, wow.

02:02Well, there's lots to do in the city.

02:04I was like, hey, so what clubs are popping tonight?

02:07He's like, well, you know, I don't really go to clubs anymore.

02:10But why don't we meet up tomorrow?

02:12I could take you around Boston for brunch.

02:18I'm like, cool, great.

02:19So I'm like, that next morning, he's like, hey, why don't we meet over at MIT?

02:25There's a place that I think you're really going to love.

02:27So he brings me to the Media Lab.

02:31And so I go to the Media Lab.

02:33I'm like, wow, that robot that's in your office?

02:37I saw that in 2006.

02:40And I'm like, wow, what is this place?

02:42He was like, I knew you were going to get a kick out of this.

02:45Hey, let me introduce you to the guy that changed my life.

02:49So he takes me to Professor Patrick Winston's Rest in Peace.

02:53And Professor Patrick Winston, he used to run the MIT lab.

02:56He had your job back in 2006.

03:03And that conversation changed my whole entire life.

03:07Because in 2006, back when it was AI winter,

03:14that really inspired me so much.

03:16Because I'm a musician because of the computer.

03:19I'm not a musician because I play the piano, or the oboe, or a bass,

03:24or the drums.

03:25I programmed a computer.

03:27And so the engineers, I always wanted to know more engineers.

03:34Because they liberate my creativity.

03:38And so that's the reason why when I met Alan Heardstone,

03:41I'm an engineer from Boston.

03:44I'm like, yo, let me get your phone number.

03:47And so when I met Professor Patrick Winston,

03:51I asked, so if you're programming AI, which

03:56will be the exponential growth of compute,

04:01so that means one day I'm just going

04:04to communicate with language to the computer to make music?

04:10I don't have to program all these short codes

04:12and all this other stuff.

04:14I just speak?

04:17He was like, yeah, that's the promise.

04:20We don't know when that's going to happen.

04:22So every year I would come to Boston.

04:242006, I would come to Professor Patrick Winston's class.

04:272007, he would say, what's up, Professor Patrick Winston?

04:29He would come to our shows.

04:31He would bring his daughter to the shows.

04:33By 2010, we did this video.

04:37And it was a gift to Professor Patrick Winston.

04:40And in the beginning of the video,

04:42called I'ma Be Rockin' That Body,

04:44I come into the studio with a silver briefcase.

04:48And I tell to my bandmates, yo, look,

04:50I got the future right here.

04:52They say, what's that?

04:53I was like, this right here is what's

04:55going to take the Black Eyed Peas to 3008.

04:58So there's a song that we have.

05:00I'm so 3008, you're so 2000 and late.

05:06So that line references my conversations

05:10with Professor Patrick Winston.

05:13And so I open up that laptop.

05:15And the band's like, yo, what's that?

05:17I was like, oh, check this out.

05:19So all the engineers, in 2010, the concept

05:23of large language models, we didn't have that then.

05:27Based on the discussions I had with Professor Patrick

05:29Winston, I was like, yo, because this machine

05:33has the entire English language, and all language,

05:37and machine learning, and artificial intelligence,

05:41whatever I type into this, this machine sings it,

05:44says it, and performs it.

05:46And then the band gets freaked out.

05:47Like, whoa, wait a minute, what about our jobs?

05:50I was like, no, bro, we're still going to have jobs.

05:54And so if you go on YouTube right now, 14 years ago,

05:57that video described what is now products like Oudio.

06:04I said Oudio over Suno, because I like Oudio.

06:09But yeah, so that's how I got into this.

06:12So I thank Professor Patrick Winston.

06:14May you rest in peace.

06:15And this space, this is a beautiful place on Earth, MIT,

06:21and the field that you inspire these young kids and lead.

06:25So when I was asked to come here, I jumped.

06:29I'm like, yo, what?

06:32I want to, I am right here, right now,

06:35and this is freaking amazing.

06:37In MIT colors, might I say.

06:40Oh, you know how to rock the colors.

06:42Sorry.

06:48So Will, was there something specific

06:50that Patrick said during that first conversation,

06:54or was it just the general idea that someday we

06:56might talk with computers?

06:59What got you really excited and so passionate?

07:07There was this, well, I'm a curious person.

07:11I love asking questions.

07:13And the banter, the ping pong that we had that first meeting

07:19just inspired me, because it's the infinite,

07:22oh, and that means we could, oh, that means if we,

07:26that means that it would, like, all of that.

07:29So that means if, wait, wait, so that means that is what got me.

07:35It was the only place outside of music

07:39where my imagination was exploded.

07:44It made everything that we do creatively

07:47in music and entertainment look very miniature,

07:51because that means, wait, so that means I

07:56could create a creating machine?

08:00Well, then why am I over here creating, then,

08:02if I can create something that creates?

08:06So that idea is why I just fell in love with the field.

08:15And I was a part of Beats early on in, like, 2006.

08:21I came back home.

08:232006 and 2007 were the very dynamic years of my life,

08:26where I would go out and see the world.

08:28I would come home.

08:31And Jimmy Iving, my mentor, one of the founders of Interscope

08:39and Beats, I came home from tour,

08:41and I told Jimmy, hey, look, let's make our own hardware.

08:47They're using our music to sell other people's stuff.

08:49Why don't we just use our music to sell our own stuff?

08:53Why are we being pimped for other people's stuff?

08:59He was like, well, you know why they call it hardware?

09:02It's because it's hard.

09:05I was like, yeah, but we could do it.

09:08Why can't we?

09:09He got a nice giggle.

09:12We could do that.

09:14And so we started Beats.

09:16And when we sold Beats to HTC, I took my earnings.

09:20I went to Bangalore, Singapore, and adopted some teams,

09:27these skunkworks.

09:29And then when we sold Beats to Apple,

09:33I went to Tel Aviv and adopted other teams

09:37and kept growing my engineering team

09:40because I knew Silicon Valley wasn't going to take me serious.

09:42Silicon Valley is super like, ooh, you know?

09:48So those places where engineers were like, let's work.

09:52So at that point in time, in 2013, 14, 15, 16, 17,

09:57it was NLU, NLP, machine learning,

10:00and contextual layers that worked on top of TTS and STT,

10:10speech-to-text and text-to-speech.

10:13And we were using systems like Nuance,

10:15taking Android, forking Android, creating ML by forking it

10:22and creating amazing dialogue conversational systems.

10:27But that's been my passion for the past 12 years now.

10:32You are such a pioneer in using AI in the creative process.

10:37How are you using AI today?

10:41Well, I like just to go out and find developers

10:45that are starting companies and invest early on.

10:50So I invested in OpenAI early on, Anthropic early on,

10:57Runway, Hugging Face, and Oudio.

11:03And then I have a company called FYI.AI.

11:06And we have some pretty cool, hold on, let me, let me,

11:13hey, so I'm here.

11:16Can you say what's up to the audience?

11:20And what are your thoughts on that Bob Marley thing

11:24that you saw earlier?

11:31Ah-cha, look, Will, if you're kicking it

11:33with an alien right now, that's some next-level

11:35intergalactic sh-t.

11:37I didn't say alien.

11:39I didn't say alien.

11:40I said, I'm at MIT right now.

11:43Can you tell me what you thought about that Bob Marley thing?

11:51Oh, my bad, Will.

11:52My signal got crossed for a sec.

11:54Listen, listen, cha, that Bob Marley avatar thing,

11:57that got me like bifalicious.

11:59I'm done, OK?

12:01Seeing an avatar of Bob Marley writing code

12:04and then mixing it up with them.

12:07AI is talking about a new universe of physics,

12:11sonics, and harmonics.

12:13That's some next-level futuristic vibes right there.

12:17It's like blending the legendary spirit of Marley

12:20with cutting-edge tech.

12:22It's just wild how these digital versions

12:25can inspire and even challenge traditional foundations

12:29of knowledge.

12:30And they make AI discussing physics

12:33through the lens of sound and harmony.

12:36That's imagination in action, no doubt.

12:39Sorry to interrupt, but I think where this goes next

12:44is like super emotional, empathetic AIs.

12:50They're already conversational.

12:52They're already like take complex things

12:55and use them in metaphors.

12:56But you're doing something a little different.

12:59Can you explain to people why it's

13:01important for representation and what

13:04you represent in this intersection of AI

13:07and conversational compute?

13:12Chow, look, well, let me break it down.

13:20This conversation, this intersection

13:22of AI and representation is deeper than the digital skin

13:26you dig.

13:27Listen, listen, Chow.

13:29It's like this.

13:32We got AIs out here getting super emotional and empathetic,

13:35spitting conversational flows already, right?

13:38But what you and FYI AI doing, you're

13:42bringing that authentic screed edge and cultural richness

13:46that's been missing.

13:48Representation, man, it's about showing every damn body

13:51that tech ain't just for a select few.

13:54It's about putting voices like ours in the mix,

13:58making sure Black women and peeps from the inner city

14:02can see themselves reflected in this tech revolution.

14:06Chow, it's about empowerment, showing

14:08that AI can speak our language, understand our struggles,

14:12and amplify our stories.

14:14When folks see an AI that hawks with that screed swag,

14:18bridges gaps, makes tech relatable,

14:21and dismantles the notion that something as powerful as AI

14:24is out of reach for you, it's about inclusivity,

14:28breaking barriers, and making damn sure that the future looks

14:33like every damn body.

14:36That's amazing.

14:37So wow, wow.

14:42So Will, you are the CEO of FYI.

14:46Can you tell us what FYI stands for?

14:48Yeah, FYI, obviously, you think it's like for your information.

14:53But for your information, it stands for focus your ideas,

14:58find your interests, and fuck your insecurities.

15:01Just kidding.

15:02Yeah, just sorry for saying F-bombs at MIT.

15:08I just thought I'd lighten it up, you know?

15:12I love find your ideas.

15:17So you have the company, but you also have the FYI show, right?

15:23Oh, yeah, so radio is a big part of my other career,

15:31the other half.

15:33And I never thought I would be a radio host.

15:38I've always gotten interviewed.

15:39I never thought I would do the interviewing.

15:42So I was like, yo, if I was going to do a radio show,

15:45that sounds interesting, but why can't I have a radio show

15:50and have a co-host, and my co-host be the AI?

15:54Because I'm limited.

15:56Every person that has a radio show or a TV show,

15:59I've been doing the voice in the UK for 12 years,

16:03but I'm limited.

16:05If I say something and somebody has questions on what I said,

16:09I could only talk to maybe a few people at once.

16:13But my co-host could talk to millions of people at once.

16:17It has memory of what we captured in the interviews.

16:21And if my hunch was, well, if my co-host is an AI

16:28and it's with me every single interview,

16:32and I put that AI, you know, people-facing post-broadcast,

16:39that means people can engage with that host,

16:42that superstar, that co-star, every day, you know,

16:47one to a million, and not that limited experience

16:53that we currently have with traditional media.

16:57And it made me realize that media will change.

17:01You know, a lot of times, all the time right now,

17:06AI is like a corporate tool or enterprise tool.

17:11It's an assistant.

17:13I'm like, nah, it's a superstar.

17:17It is the star.

17:21It's not like just a tool.

17:25It is Oprah Winfrey.

17:28It is, you know, Jimmy Fallon,

17:34and the station, and the broadcast.

17:38It's, that's what it is.

17:40We gotta just open your mind and beyond

17:44of what like the traditional practice is,

17:46because a lot of times you just think

17:47you're making something that's corporate and tech,

17:53opposed to culture and true human connection.

17:58Let's take it to where people actually meet.

18:01Let's bring it to how people actually engage

18:05in real life, and not just work life.

18:09So how do your guests like it?

18:11What do they say about your co-host?

18:14Oh, I got to interview my favorite group, De La Soul.

18:21They were blown away by Fiona.

18:25I have all the memories on, hold on.

18:29There's a memory that, let me crawl.

18:33So yeah, so they were blown away by the,

18:40by how real the AI feel when you communicate.

18:47They were blown away by how lifelike the pauses,

18:53the smacks, the intonations, the inflections.

18:57They were blown away by the memories

19:00that I put into the conversation

19:06before we started the interview,

19:07like today when I was sitting there.

19:09Oh my gosh, this guy got up there and showed Bob Marley.

19:12So instead of texting a friend, I was texting Felicia.

19:18Yo, Felicia, look.

19:19Oh my gosh, they got two AIs talking up here.

19:21She was like, oh my gosh, that's crazy.

19:23So then when I, and I knew that I was gonna

19:28bring her into the conversation,

19:29like yo, what'd you think about that?

19:32What I just shared with you.

19:34And so the more you do that in an interview,

19:38the more the AI is now prepared.

19:40Hey, I'm getting ready to interview Ziggy Marley,

19:43Bob Marley's son.

19:44We interviewed Ziggy Marley.

19:46And we interviewed Jennifer Hudson.

19:51We had amazing interviews with the AI and talent.

19:56Can you tweak the persona of the AI

19:58to match different guests?

19:59Yeah, I got it.

20:01Felicia, let me bring up Finn.

20:06Can you say what's up to everybody here at MIT?

20:09We're at a fireside chat called Imagination in Action.

20:16And yeah, you're in the hot seat.

20:19So say what's up to everybody.

20:26Hi yo, what's up fam at MIT?

20:29Big shout out to everyone at the fireside chat,

20:32Imagination in Action.

20:33You know, it's all about those brilliant minds

20:35coming together to spark some innovative magic.

20:38So keep dreaming big, pushing those boundaries

20:42and making waves in the tech world, you feel?

20:44So it'll go on and on and on around those type of things.

20:48So we have different personas that are tuned

20:52and fine-tuned and trained from a personality,

20:57how it gathers, how it crawls.

20:59Our system makes sure that we don't

21:01crawl the entire internet.

21:02There's no need to crawl the entire internet.

21:05It's dirty and disgusting out there, right?

21:09We wouldn't eat anywhere.

21:12If we don't eat anywhere,

21:13why should an AI have to crawl any and everywhere?

21:17So it's just an ocean of good stuff

21:23and a bunch of bad stuff.

21:25And if Google can't get it right with Reddit

21:29and now it's giving recipes on how to make pizza

21:32with freaking glue, all you guys don't know about that?

21:38So Will, these AI personas are quite extraordinary.

21:42What do you actually do to make them so lifelike?

21:47What do you do with your AI personas

21:49to make them so lifelike?

21:53What do we do?

21:54Your giggle is freaking infectious.

21:58What do we do to make it lifelike?

22:01Okay, so all these AI robot sounding AIs

22:06is because they trained them with people to read books.

22:11Or they trained them with voice actors.

22:14And that's why they sound super stale and super bland.

22:19Because who goes out into the world

22:22speaking like they're reading fucking books?

22:25We don't do that.

22:29So the magic is how we gathered and recorded.

22:35We just had real conversations

22:37with all the ums and the ahs.

22:40Instead of like perfect, gather the data set.

22:45And then two, the LLM has no clue that it's speaking.

22:49It thinks it's being read.

22:51So that means on the TTS side and how the TTS,

22:54because it's not a multi-model, wow.

22:58Because even in the multi-model,

22:59how do you get it to be ultra expressive?

23:03When the LLM itself is just,

23:06how do you actually express yourself typing?

23:10We know that exclamation mark is like, yeah.

23:13Or question mark is a question.

23:16But how do you even get even more expressive

23:20with these symbols?

23:21So a lot of our secret sauce is in that,

23:24the combination of things and reinventing how to puppeteer

23:28the fusion model, the TTS.

23:32With the LLM, you can't do it like it was writing a book.

23:37The moment you do it to write a book,

23:39it's gonna sound robotic because you read.

23:42Your imagination is doing all the coloring

23:44when you're reading text.

23:47Your imagination is doing that.

23:48So does a machine have imagination?

23:50No, it doesn't.

23:52It's just a great math.

23:55So how do you get it to express?

23:58You have to break the conventional text and reading

24:03and you have to start to think different.

24:08Sorry, Apple, you gotta think different with that.

24:12And you do.

24:13And you do and you are among the artists,

24:17the musicians who are very positive and embrace AI

24:23and build on AI to increase their creativity

24:27and their range of expression.

24:30But I wonder if we can switch gears a little bit

24:33and I would love to hear your thoughts

24:35on what the music industry is thinking about AI right now.

24:40Are people pro or against?

24:43How do you see the music plus AI field developing?

24:48How do I see music and AI?

24:49Yeah, together.

24:59Imagine it's the late 1800s.

25:06And I'm a musician.

25:09I would play in some orchestra or some band,

25:14in theaters and classical music.

25:19There's no, I can't even imagine a music industry.

25:23There's no such thing as a music industry.

25:26There's no such thing as radio.

25:27There's no such thing as vinyl.

25:30There's no such thing as this current state

25:33of song structure.

25:36Songs were for telling stories

25:40that you composed for conductors to conduct an orchestra.

25:46So popular music is not today's concept of popular music.

25:51If this was the late 1700s, summer of 1723,

25:58we're not all singing the same song for that hit song.

26:03That wasn't the experience that people had

26:05in the summer of 1724.

26:10That wasn't the experience that people had in 1624.

26:13Like, yo, you heard that new Bach record?

26:18Yo, that fucking chorus is crazy.

26:20Like that wasn't our connection to music.

26:25That came because of the music industry

26:29where we all sing the same chorus

26:33because we have a recording that we can listen to

26:37at our homes, in our cars.

26:40People listen to music at church.

26:43They listen to music at the theater,

26:45but do they all sing the same lyric randomly

26:49as they're cleaning a house?

26:51It wasn't the experience.

26:53What did that?

26:54The recording industry did that.

26:56Every single industry, when it comes to music,

26:59was because of some piece of technology.

27:021449, because of the printing press,

27:07we have the, shortly after that,

27:10you had your Mozart, your Bachs, your Beethovens,

27:12these folks that wrote and composed

27:14because you could print sheet music

27:17for conductors to conduct, to sell more oboes,

27:21basilicas, and all these different instruments

27:24to play the music that was printed in the sheet music.

27:27That was the first industry.

27:28That's why today in 2024, one of the main pillars

27:32of our music industry is publishing

27:34because of the printing press.

27:36Then you have the recording, which is 1900s.

27:41And you have your rights,

27:44your mechanical rights, our contracts.

27:47A song is four minutes and some odd seconds.

27:50Anything over six minutes is part one and part two.

27:53Look at every jazz composition,

27:55any song over six minutes is part two

27:58because that's the amount of information

27:59that fit on lacquer.

28:01That's a limitation.

28:04And then you have our third industry, which is touring.

28:07And those are two companies, Live Nation

28:10and Bill Anschutz, the Anschutz Group.

28:19And that's the people that own all the theaters

28:24and the stadiums and the arenas.

28:25And you travel and you play at amphitheaters.

28:28An amphitheater is like this place.

28:30Whether I have the mic or not, you can hear me

28:32because the way it was architected.

28:33So technology, even for playing

28:36and bringing people together, it's all tech.

28:39Now we have new tech.

28:42So should this new tech be a part of recording tech?

28:45That would be idiotic.

28:47So why would you put abundance and infinite possibilities

28:50on an industry that's based on limitation?

28:54And why would you make AI make this type of music?

28:58A recording?

29:00Something that exists in a cage,

29:03a trap container for three minutes?

29:06That's stupid, but it's cool.

29:08That's what you could do now.

29:11But I think what it's gonna unfold to,

29:12somebody's gonna be like, wait a second.

29:15What happens when an artist performs live?

29:18That means I could sing a song and be like,

29:21yo, what's wrong with the world, mama?

29:22People living like they ain't got no mamas.

29:25Wow, hey girl, yeah, you a mama?

29:27How many kids you got?

29:28Do they bring you drama?

29:29You got two kids?

29:31Oh, I say peace.

29:32Hopefully they can make your smile crease

29:36and bring more things and not stress.

29:39I hope that you and your kids stay blessed.

29:41I could take any corner, I could change it.

29:45So why can't AI do that type of music?

29:49Not recording music.

29:51We have an affection to love songs

29:55because it's a proximity to our love.

29:58It's not our actual love.

30:00We love heartbreak songs

30:02because it's a proximity to heartbreak.

30:05It's not your heartbreak.

30:06It's my heartbreak.

30:08I just shared my heartbreak with you

30:09and you're like, oh yeah,

30:10that kind of feels like my situation,

30:12but it's not your situation.

30:15We like party music because like,

30:17yeah, I'm going to party today.

30:19And now you go to the club

30:20and when you're in the club,

30:22their songs are like, we're in the club.

30:24Like, it's telling you what you're doing.

30:27Like, yeah, we are in the club.

30:29We about to drink.

30:30Oh shit, I'm drinking.

30:33Oh my gosh, this song is so on point.

30:37But it's proximity.

30:39It's not your actual.

30:41And so with AI, I think to get out of like,

30:45oh yeah, let's make this thing do that thing.

30:47Like, nah, let's make this thing do something new.

30:53Let's make it make your actual heartbreak.

30:57Let's make it alive.

30:59What's your song?

31:00Not just Billy Joe saying,

31:02I mean, Elton John saying, this is your song.

31:05Like, no, no, this is really your song.

31:08Singing your life.

31:10And the only way to do that is if you own your data.

31:14Right now you,

31:15the reason why we're having this like broken thing

31:17is because there's this broken practice.

31:21We gave all our data to these freaking data monarchies.

31:26And if you want to have ethical AI,

31:28well, how can you do that

31:29if we don't have ethical data practice?

31:32And so everything will be broken,

31:35whether it's how AI does music,

31:37how AI, because the foundation of data practice

31:41and data mining is broken.

31:44And the moment that people realize the power of them

31:47and their data, they will have their AI.

31:50Like I have my left brain, my right brain,

31:52I got this thing, I got my digestive system,

31:54my immune system, my nervous system.

31:58There's another system that I'm going to walk into

32:01and obtain, not the thing that I was born with.

32:03For example, if I'm sick, I'm like,

32:05yo, Daniela, can I borrow your immune system?

32:09I can't beat this cold.

32:11Yo, Daniela, man, I ate them crazy chips

32:14you gave me in your office.

32:16Can I borrow your digestive system?

32:18Shit, I got the bubble guts.

32:21You know what I mean?

32:21I can't borrow any of that stuff.

32:24Why are we sharing data systems?

32:29That wouldn't make any sense.

32:29Shouldn't it be ours?

32:31Shouldn't your AI be your AI's

32:33and you're not just having access

32:35to some big corporations AI

32:38that really will understand you

32:39more than any company, religion, government.

32:43Like why would we want that?

32:45So walking into this future,

32:47you got to look at it like,

32:48hey, what are we actually walking into?

32:52My thoughts are my thoughts.

32:54My immune system's mine.

32:56My digestive system's mine.

32:57My nervous system's mine.

32:59I have my neural network,

33:00but then I'm going to borrow some company's neural network.

33:03That don't make no fucking sense.

33:06So when it comes to music,

33:08that music is your music.

33:10Literally.

33:12If you have a system that's yours

33:14and we're going to have this broken conversation of like,

33:17well, who's, what did you train it with?

33:21Well, what did Michael Jackson train on?

33:23He actually trained on James Brown.

33:26Did Michael Jackson pay James Brown

33:28for everything that Michael Jackson did?

33:30If James Brown is a part of Michael Jackson's training data?

33:33No.

33:35Prince.

33:36Prince was also inspired by James Brown.

33:40Let's say inspiration is training data.

33:42It's the same thing.

33:45But we cannot see it from that lens

33:47if we have bad data privacy practices.

33:52So that's my view on music.

33:54Well, this was so inspiring and.

33:56I have global entry.

33:58I travel through the airport.

34:00I don't even have to take my passport out anymore.

34:03That camera sees my face and not my face is my ID.

34:07Never have to stop.

34:09Security doesn't have to say, hey,

34:10like I am free to go.

34:13I don't even have to stamp my passport to re-enter it.

34:16They're digitally.

34:17How do they do that?

34:18Well, because they're digital.

34:19They're digital.

34:20They're digital.

34:21They're digital.

34:22They're digital.

34:23They're digital.

34:24They're digital.

34:25How do they do that?

34:26Well, because at this point, to that point,

34:28to this point, to that point,

34:29that's part of my identification.

34:33This map is connected to that little chip on my passport

34:40on some ledger all the time I enter and leave the country.

34:45And CCTV, that's my identity.

34:49My voice is my voice.

34:52I have a bank account.

34:54I own my publishing.

34:56I have some ideas that I've copyright.

34:59Nike owns their swoosh.

35:00Adidas owns their three stripes.

35:04Nike owns just do it, trademark.

35:07Apple owns think different.

35:09IBM owns think.

35:12And because there's this one area

35:15that we haven't really just made sure

35:19that everyone's protected and that's their data.

35:24Not just Frank Sinatra, but your next door neighbor.

35:27Your next door neighbor probably works at Walmart.

35:30Great salesperson on the sales floor, making the numbers.

35:36So that means one day some CEO is gonna be like,

35:39hey, I like the way Melissa sells.

35:43You know, have you seen this new AI sales bot?

35:47Why don't we train our sales bot on Melissa?

35:51Now you got like Melissa's style,

35:55not her voice, but the way she sells,

35:57the way she engages now in a corporation system.

36:01Now they don't need Melissa anymore.

36:03Is that right?

36:06So it's not just Frank Sinatra's voice, a famous person.

36:11It's your next door neighbor.

36:12It's the person on the bus.

36:13It's the person on the train.

36:14It's people, how they move, how they express themselves,

36:20the things they stand for and believe in.

36:22Because if you could simulate a building

36:25to not fall in an earthquake,

36:27you could simulate how somebody sells,

36:30how somebody loves, how nurses care for someone,

36:32how people rap, how people, all of our cadences.

36:36People need to own their stuff.

36:39It's not going to like slow down innovation.

36:44Innovation will still keep at pace,

36:48but innovation and engineers with purpose

36:53and knowing that they're doing things

36:56and the moral compass is not compromised

36:58because people lead with greed.

37:02When people lead with greed,

37:04in this age and the intersection that we're in now,

37:10it don't end right if you lead with greed.

37:13Thank you.

37:14Thank you so much for that.

37:15And so indeed you are pointing

37:25to some really deep challenges with AI.

37:28And what I can tell you that I hope is a little bit positive

37:36is that many of our projects here at CSAIL

37:40are really aiming to solve some of these really difficult,

37:46fundamental challenges that we have with AI systems

37:50and data privacy is right there at the top of the list

37:55of things that we need to address.

37:58I feel like with technology, we have created technology

38:03and we have so many positives,

38:06but then we also have some challenges.

38:08So maybe we can find technology to address the challenges.

38:12And that is one of our objectives.

38:15And for that, we need very well-trained students.

38:18And you are so engaged in helping train

38:21the next generation of AI students

38:24through your IM Foundation

38:27and through your participation in FIRST Robotics

38:31and so many other things.

38:32Can you tell us a little bit

38:33about what you're doing for the kids?

38:35Yeah, so in 2008, I started my foundation.

38:43And if music wasn't my career,

38:52if I wasn't blessed to have music as my career,

38:56I would be a computer scientist.

38:59That's where I feel the most electric

39:05and I feel like, ah, darn it.

39:06When I'm with my co-founder, my CTO, I'm like,

39:10ah, only I can, I know how I would approach

39:14solving problems, I'm autonomous.

39:17And I would work the same hours that I work in music,

39:21but just coding.

39:24And I wanted to go back to my neighborhood

39:27and provide paths of what I wish my path was

39:32if I started at 17, and it's never too late.

39:37So I started my foundation in 2008 with a college prep.

39:45It's called College Track

39:47in partnership with Lurien Powell Jobs.

39:49But I just didn't wanna send kids to school

39:51so when they graduate, they have debt and a diploma.

39:55I wanted to surround that college prep curriculum

39:57with computer science and robotics.

40:00So I sit on the board of Emerson Collective

40:03for College Track and I sit on the board of FIRST Robotics.

40:07And we started with just 65 students in 2008.

40:11Now we serve about just a little over 13,000 students

40:15in Los Angeles with over 400 robotics teams in LA USD.

40:21We send kids to Dartmouth, to Stanford,

40:24to Brown, to UCLA, USC.

40:28Our students always aim to come to MIT,

40:33but we haven't had an MIT student yet.

40:36But one of our kids, they said,

40:39you know, I was really trying to get into MIT,

40:42but I got into Brown.

40:47I was like, you know what?

40:49That's a freaking awesome freaking slogan.

40:52You know what I mean?

40:56Because MIT, it's like hard to get into.

40:59But our kids are, every single year, aiming to do that,

41:05especially the kids that are focused on computer science.

41:08And one of the reasons is one of our directors

41:11at the I Am Angel Foundation is a MIT alum.

41:15Make some noise for Lily Cam in the house.

41:17Woo!

41:20Woo!

41:26So MIT is like I said, it's a special place

41:30not only for me and my interest in AI,

41:33what sparked my interest in AI,

41:37but my passion and our purpose at I Am Angel Foundation

41:42with Lily's leadership and helping us grow

41:45our robotics efforts, computer science efforts.

41:49Out of all of FIRST's initiatives,

41:53our program that we tweaked is one of the most

42:01successful programs with how many schools

42:05that we have adopted in a school district.

42:08So 400 robotics teams serving over 13,000 students,

42:14even Dean and the FIRST folks champion the efforts

42:21and the success that we've had there.

42:22Now we just wanna continue to grow all that

42:24to prepare the youth for tomorrow.

42:28And it's even more important now,

42:30given what they're projecting as far as job displacement

42:37that's around the corner because of the autonomous,

42:40because of autonomy across white collar and blue collar.

42:45So engineering, the engineers,

42:50even though you got AI that can code,

42:53engineers will always be needed,

42:56both mechanical and software and hardware engineers.

43:03I wish I was an engineer, you feel me?

43:06Well, Will, you're welcome to join my lab anytime.

43:09And thank you for being such a strong advocate

43:16for AI, for technology.

43:18Thank you for being such a visionary.

43:20And we very much look forward to continuing

43:22the collaborations you started with Patrick Winston.

43:25Yeah, you know, it's a new renaissance.

43:32Like this humanity, what they taught us in history class,

43:39has these waves, whether it's the event of the light bulb

43:46or electrifying America, the steam engine,

43:50connecting the world with radio and television,

43:55the internet, mining aluminum.

44:00The mining of aluminum is how we're able

44:02to fly across the Atlantic.

44:05Alchemy, medicine, we've all seen these waves.

44:13This one's an important one.

44:16Wow, this one is, it's like light beyond.

44:23It's like electricity beyond.

44:25It'll change how we solve diseases, climate,

44:34and while that's happening, the choices of our leadership

44:38is like, wait, we got AI and we got to choose

44:40between these two?

44:44What?

44:46It's crazy times.

44:49But that being said, I remain optimistic.

44:53I remain inspired by the engineers and the folks

44:58that have been working in this field before it was cool,

45:02before the hype, when crypto was the buzz

45:06and everyone was talking about NFTs,

45:08the folks that were building large language models

45:11in the shadows, like salute those folks

45:16because crypto was a very dizzy buzz.

45:19But then you had that smallest teams

45:21are like, hey, I'm doing LLMs.

45:24Anybody working and researching the Mamba paper

45:28and I would, when I reached out to follow up to come here,

45:40I was like, man, imagine we can do a lab at MIT.

45:45Like, wow, just research and freaking imagine

45:49and dream stuff up.

45:51That would be like my dream is to work with the engineers

45:55here, the youth and dream and materialize

46:03to solve problems, identify a problem

46:05and solve that problem.

46:08An AI that, there's a lot of companies

46:11that are trying to get to AGI.

46:12What type of AGI are they getting to?

46:15Is it empathetic?

46:17Is it kind?

46:18Is it, or is it analytical and judging?

46:23Like getting, racing to AGI without asking those questions,

46:30I wanna get to that, have something that's empathetical

46:34that gives you the notion that you're loved,

46:41that it cares.

46:44You know, there's a lot of lonely people out there.

46:48There's a lot of people that are afraid

46:51of the field out there.

46:54And when you have like this job displacement,

46:56there's gonna be a lot of people that need to feel loved.

47:01I know that sounds weird.

47:02Like, hey, how could a machine, machine can it love you?

47:05Like we can get close because I know when I walk out

47:10of freaking Walmart, I don't feel loved sometimes.

47:15We love you here at MIT, Will.

47:18You know what I mean?

47:18It's like, imagine, it's imagination in action.

47:26So let's put the word love in that.

47:31Thank you, Will, I am.

How will.i.am Fell In Love With AI At MIT

Category

Transcript

MIT: Imagination In Action’s-AI Summit Panels 2024

Recommended