A serendipitous meeting while on tour in Boston led will.i.am to MIT's Media Lab, and established a lifelong passion for artificial intelligence.
At Imagination In Action's 'AI Frontiers and Implications' event at MIT CSAIL, will.i.am sat for an interview with Daniela Rus, Director of MIT CSAIL.
Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1
Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:
https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript
Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com
Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.
At Imagination In Action's 'AI Frontiers and Implications' event at MIT CSAIL, will.i.am sat for an interview with Daniela Rus, Director of MIT CSAIL.
Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1
Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:
https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript
Stay Connected
Forbes newsletters: https://newsletters.editorial.forbes.com
Forbes on Facebook: http://fb.com/forbes
Forbes Video on Twitter: http://www.twitter.com/forbes
Forbes Video on Instagram: http://instagram.com/forbes
More From Forbes: http://forbes.com
Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.
Category
🤖
TechTranscript
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.