Optimism for AI – Leading with empathy

Summary

This discussion features will.i.am, the musician and tech entrepreneur, speaking with Nicholas Thompson from The Atlantic about AI’s role in creativity, education, and society. The conversation takes place at a UN AI for Good event where will.i.am serves as a special ambassador for the AI Skills Coalition. When asked about his previous advice that aspiring musicians should study AI, will.i.am explains that while AI can help materialize musical ideas quickly, traditional skills like playing instruments and writing remain important for personal expression and mental flexibility. He describes his creative process as using AI to generate variations of his original ideas rather than relying on it for initial inspiration, emphasizing that humans should remain “the dreamers of the dream” while AI serves as an “infinite generator” of existing concepts.

The discussion reveals that will.i.am remarkably predicted current AI capabilities in a music video from 15 years ago, forecasting voice cloning and language models. In education, he focuses on teaching young people to be “audacious, ambitious, and competitive” as core skills for the AI era. The conversation shifts to AI governance, where will.i.am advocates for an “AI constitution” and licensing requirements for AI developers, similar to driver’s licenses for vehicle operators. He criticizes the current lack of oversight, arguing that society cannot achieve “AI for good” with poor data practices inherited from social media. While acknowledging good actors in the AI field like Dario Amodei and Demis Hassabis, will.i.am warns against leading with greed over human welfare. The discussion concludes with his call for innovative leadership that balances technological advancement with safety constraints, drawing parallels between AI governance and humanity’s historical success in taming wild animals while maintaining safety protocols.

Keypoints

**Major Discussion Points:**

– **AI’s role in creative processes and education**: Will.i.am discusses how AI should be used as a tool to amplify human creativity rather than replace it, emphasizing that humans should remain the “ideators” while AI serves as an “infinite generator” to elaborate on original human ideas.

– **The need for AI regulation and licensing**: A key focus on the absence of proper oversight in AI development, with Will.i.am advocating for licensing requirements for AI developers (similar to driver’s licenses) and the creation of an “AI constitution” to establish ethical guidelines.

– **Greed versus empathy in AI development**: Discussion of whether current AI development is driven primarily by profit motives or genuine concern for humanity, with Will.i.am identifying some “good guys” in the field while expressing concern about corporate priorities.

– **Data rights and ethical AI practices**: Emphasis on the connection between poor data practices from the social media era and the potential for harmful AI, stressing that “AI for good” cannot exist without proper data governance.

– **Historical parallels and societal responsibility**: Will.i.am draws connections between current AI development and historical myths/folklore (genies in bottles, “abracadabra”), questioning whether humanity has faced similar challenges before and emphasizing the need for collective responsibility in managing this technology.

**Overall Purpose:**

The discussion aims to explore the intersection of AI technology with creativity, education, and society, while advocating for responsible AI development that prioritizes human welfare over profit. It serves as both a philosophical examination of AI’s role in human creativity and a call to action for better governance and ethical practices in AI deployment.

**Overall Tone:**

The conversation maintains an engaging and thoughtful tone throughout, with Will.i.am bringing both philosophical depth and practical insights to complex AI topics. The tone is generally optimistic about AI’s potential while remaining cautiously critical of current practices. There’s an informal, conversational quality with moments of humor and personal anecdotes, but the underlying message carries serious urgency about the need for responsible AI development. The tone becomes more emphatic and passionate when discussing regulation and ethical concerns, particularly toward the end of the discussion.

Speakers

– **will.i.am**: Musician, educator, AI ambassador for AI Skills Coalition, runs robotics computer science college prep programs

– **Doreen Bogdan Martin**: Appears to be in a leadership role introducing speakers and initiatives related to AI Skills Coalition

– **Nicholas Thompson**: Journalist from The Atlantic, serving as conversation moderator/interviewer

Additional speakers:

– **LJ**: Briefly mentioned as handing over to others, role/title not specified

# Discussion Report: AI, Creativity, and Governance – A Conversation with will.i.am

## Introduction

This discussion, moderated by Nicholas Thompson from The Atlantic, featured will.i.am—musician and special ambassador for the AI Skills Coalition—at a UN AI for Good event. Doreen Bogdan Martin introduced the session, highlighting will.i.am’s role with the AI Skills Coalition. The conversation explored AI’s role in creativity, education, and governance, covering both practical applications and broader societal concerns.

## AI and Music Creation

will.i.am described his approach to using AI in music creation, emphasizing that humans should maintain control over original ideas while AI handles elaboration. He explained his creative process: “I’m sitting in Paris and I have an idea and I want to materialize that idea quickly. So I go to AI and I say, ‘Give me variations of this idea that I have.'”

He used a compelling metaphor to illustrate the human-AI relationship: “AI now has this metaphor. AI now does Pilates. AI now stretches. That doesn’t mean you don’t have to stretch your damn self. You still got to stretch. I don’t care how dope AI is at stretching and doing Pilates. You still have to be flexible yourself.”

will.i.am emphasized the importance of maintaining human creativity and traditional skills: “We are the ideators. It is an infinite generator. At this point in time, it’s just regurgitating imagination. It’s a regurgitation of imagination. We are the dreamers of the dream, Willy Wonka.” He stressed that humans should not give machines “the power to imagine first” and that traditional skills like playing instruments remain crucial for personal expression.

## Education and Youth Preparation

will.i.am discussed his educational initiatives, revealing that he started robotics and computer science programs in 2008, demonstrating early recognition of AI’s future importance. He also mentioned creating a music video 15 years ago that anticipated current AI capabilities like voice cloning and language models.

His educational philosophy centers on developing core skills for the AI era. He identified three key attributes for young people: “audaciousness, ambitiousness, and competitiveness” as fundamental tools regardless of their chosen field. The discussion touched on applying the principle of human-led ideation with AI elaboration across different educational disciplines.

## AI Governance Concerns

### Licensing and Regulation

will.i.am made a strong case for AI developer licensing, drawing an analogy to vehicle operation: “Each one of us that operate machines have a licence to operate that machine… To make AI, you don’t have to take a test. There’s no licence to deploy systems that are going to reconfigure how people think.”

This comparison highlighted the regulatory gap in AI development, where systems capable of reshaping human cognition face fewer oversight requirements than physical machines.

### Constitutional Framework

will.i.am advocated for creating an “AI constitution” that would establish ethical guidelines for AI development. He suggested this framework could help identify problematic AI systems based on non-compliance with established principles.

### Greed Versus Empathy

A significant theme emerged around whether AI development prioritizes profit or human welfare. will.i.am argued that AI development should be “human first” rather than “money first.” He mentioned several AI leaders by name (Dario, Demis, Mustafa, Aiden) as “good guys” in the field but noted they operate within systems that may not always prioritize human welfare.

## Data Practices Critique

will.i.am delivered a pointed critique of current data practices and industry rhetoric: “You can’t have AI for good with poor data practice. And we are still in process of doing that poor data practice from social media. So is this just a slogan? Are these just like word logos? This sloganism?”

This comment challenged the gap between stated intentions around “AI for good” and actual practices, particularly regarding data collection and usage inherited from social media platforms.

## Historical Context

will.i.am provided interesting historical perspective, connecting current AI capabilities to ancient concepts: “Have we been here before? Or is this the first time we’ve been here? Because it kind of sounds like things that were in fantasy movies in the past… abracadabra that comes from Aramaic, which means I create as I speak. That kind of sounds like chachi piti… genies and bottles, kind of sounds like what we are now.”

He suggested that humanity has successfully managed dangerous forces before and can develop appropriate constraints for AI while maintaining innovation.

## Key Themes and Takeaways

The discussion revealed several important themes:

1. **Human Agency**: Both speakers emphasized maintaining human control over creative and intellectual processes while using AI as a supportive tool.

2. **Education Priority**: The importance of preparing young people for an AI-driven future through proper skills development was a recurring theme.

3. **Governance Urgency**: The conversation highlighted the critical nature of current decisions about AI regulation and ethical frameworks.

4. **Authentic Implementation**: will.i.am’s challenge to “AI for good” rhetoric emphasized the need for genuine reform rather than superficial slogans.

## Conclusion

The discussion demonstrated the value of bringing creative and journalistic perspectives to AI governance questions. will.i.am’s unique position as both artist and educator allowed him to bridge practical applications with broader societal concerns. His willingness to challenge industry rhetoric and call for concrete governance measures provided substantive direction for future AI development.

The conversation emphasized that while AI offers tremendous potential for creativity and education, realizing its benefits requires deliberate choices about regulation, ethics, and maintaining human agency in the process. The urgency of these decisions, combined with the practical frameworks discussed, suggests clear next steps for responsible AI development.

Doreen Bogdan Martin: A really, really exciting announcement. We have a friend joining us, a friend I met a couple of years ago. He’s everywhere. Everybody knows him. And I’m so happy to introduce him today. It is Will. I am. Will, it’s wonderful to see you again. When I met Will, we were actually with Mark, and we talked about all things tech. We talked about skilling, and we talked about young people and connecting schools. And today, we’re going to talk about how Will is our special ambassador for our AI Skills Coalition. And we’re so happy, Will, to have you with us. So ladies and gentlemen, Will, I am. Thank you very much, Doreen. That was amazing. All these initiatives, and we’re thrilled to have you along for the ride remotely, Will, I am. We’ve arranged a fantastic conversation partner in the form of the Atlantic’s Nick Thompson, who is going to be sitting in the hot seat behind us. But for now, it’s over to Nick for more. And thank you very much, Doreen. And everybody, we’ll see you soon. Thank you. Should we head off the front? Wonderful.

Nicholas Thompson: Thank you very much, LJ. Thank you very much, Doreen. Will, it is wonderful to see you. You can’t see the room, I don’t think. But just so you know, it’s full. There are people lined up. It is exactly like your concert last week in Casablanca. All right. I want to start with something you said to me a couple of years ago. And I asked you, I said, if my kids want to become musicians, what should they do? And you said they should start to study AI. And I’m curious if that advice still holds, and if the way they should study it has changed at all in the last two or three years. Wow. That was three years ago?

will.i.am: When was that? Yeah. I mean, if you want to make music, materialize ideas into audio results quickly, then you should dabble with AI. If you want to just play music, you still should learn an instrument. If you want to express yourself and wrench yourself out after you’ve been through something harsh or you’ve experienced something wonderful, it’s still an important skill to want to learn to write. But you don’t necessarily have to do that to get an idea out. You don’t have to do that. It’s great to know. But there’s this tool now that you could just explain how you feel, and it’s going to give you the proximity of materializing songs. And that’s still fulfilling for a lot of people. How do you think, for me and hyper-creatives like myself, we still want to stretch ourselves? AI now has this metaphor. AI now does Pilates. AI now stretches. That doesn’t mean you don’t have to stretch your damn self. You still got to stretch.

Nicholas Thompson: I don’t care how dope AI is at stretching and doing Pilates. You still have to be flexible yourself. How do you use AI when you’re writing music now versus how you used it a couple years ago? I should say, by the way, that Will forecast the entire future in a music video from ten years ago where he starts to talk to his bandmates about voice cloning, and then his bandmates run out and start yelling at him and go dance in the desert, which you should all watch. But explain now. It is amazing.

will.i.am: I mean, Fergie heads off on a motorcycle in anger at Will, who basically forecast Eleven Labs. Regardless, how do you use it now? It was 15 years ago. That video was 15 years ago. It forecasted not just Eleven Labs, it forecasted Suno, Udio, and LLMs, because in this sentence that I explained to the band, I’m like, look, you take the entire English language, you take my voice. So there was no fusion models, diffusion models, there was no LLM models at the time. I was paraphrasing it based on how this might come to be. But it kind of forecasted. Check it out. The song is called I’ma Be Rockin’ That Body on YouTube that kind of paints where we are. So what was your question again? Sorry, I just wanted to clear up how long ago it was. You should all watch that if you haven’t already watched it this morning. The question is how you use AI now in your creative process as a musician versus how you used it back when the tools were much weaker. When you’re writing a new song, how are you using AI? I don’t change how I write. How I write is how I write. Like I said, it’s a stretch. What I do after the stretch is different. So say, for example, I write, you know, I have an idea. And the idea is like, you know, sitting in Paris, I wish I was in Switzerland. Sitting in Paris, I wish I was in Switzerland. Sitting in Paris, I wish I was in Switzerland. Sitting in Paris… Sitting in Paris I wish I was in Switzerland. My… Damn, I’m embarrassed. I think I’m on that shit again. Sitting in Paris, I wish I was in Switzerland. Damn, I’m embarrassed. I wish… Damn, I’m embarrassed. So then I boom, I’ll form it, form it, form it, form it, form it, and I’ll keep going, keep going, keep going. And then I’ll have a form. And so that exercise of like, this stanza is shaped like this stanza, the next stanza, it rhymes with the next stanza, sitting in Paris, I wish I was in Switzerland. Damn, I’m embarrassed. I think I’m on that shit again. So those words have the same form, the same shape. And that’s cool. That’s a good exercise for me. What does that mean? It’s pattern matching. So I’m pattern matching. So this… It’s a pattern match. I matched these shapes of sounds. And that’s good. It’s a good exercise for how I see the world, because I could use that same skill to pattern match visuals. I could use that same skill to pattern match ideas and concepts. It allows me to predict what’s coming based on the shape of… of now, too. So that same shaping words allows me to shape ideas, shape how we are, shape what I see. So I don’t want to give that to the machine. So what do I do with the machine now? Well, with the machine, I can be like, OK, here’s what I have. Give me 100 of these. Because it would have took me a whole freaking day to come up with 100 variations. So I will come up with the original form and then throw it in the machine, and the machine will give me a bunch of them. And I never want to give the machine the power of imagining the first. We are the ideators. It is an infinite generator. At this point in time, it’s just regurgitating imagination. It’s a regurgitation of imagination. We are the dreamers of the dream, Willy Wonka. That’s the famous line from Willy Wonka. I forgot who was the original poet that wrote that, but that’s where Willy Wonka borrowed from. I like dreaming. I like imagining. And so, yeah, I use it to regurgitate my imagination. I don’t use it to imagine.

Nicholas Thompson: So this is a hugely profound idea, and it gets to some of the core questions about how we learn, how we’ll interact with machines. And so I want to ask you a little bit. You are, of course, an educator. You have many students. You brought many people into the education system. As you think about teaching them to learn, and as you think about how you teach them to learn in this age of extremely intelligent AI, does that principle still hold in every field? You come up with the core idea. You come up with the core concept, and then you can let the machine elaborate on it. You keep your mind sharp so that you never cognitively offload the capacity to get the original idea. Do you impress that upon the kids you work with?

will.i.am: The kids that we work with, we encourage the kids to be audacious and ambitious and compete. And those three elements, no matter what their field is, is the tools they need for tomorrow. That’s the source, the core that they need. You need to be audacious to compete, and you need to be ambitious to get up once you got knocked down. In the past, we were just doing that with humans. And now, based on that pattern matching and seeing what’s possible based on what’s happening now, what’s plausible based on what’s happening now, to be able to see like, hey, this is how we’re going to be making music in the future in 2015. Sorry, 15 years ago, in 2010, when we made that video. Just two years before that, I started my program, 2008. And that program was a robotics computer science college prep program. Because that same intuition of, ooh, maybe the future is going to look a lot like this, let me make sure kids from my neighborhood are prepared for that future. So that audaciousness and ambitiousness is what we try to mentor and encourage our kids to fuel themselves with, ambitiousness, audaciousness, and the will to compete. You’ve talked a number of times about the need for an AI constitution and the need to make sure that as we head into this era of AI, we don’t make similar mistakes that we made in the Web 2 era. I’m curious right now what you think the key principles should be in an AI constitution. Wow, you’re pulling up all the things, all these gems.

Nicholas Thompson: So yeah, so it did. You say all these smart things. I’ve been reading them for years, man. Of course I know these things. So yeah, so the AI constitution, once again, a pattern match. I’m like, why isn’t there an AI constitution? And whoever doesn’t sign it, that’s the AI you don’t want to mess with. We don’t even have that right now. And before we even get that complex, most of these people in the audience, most of you guys, you have a driver’s license. Some of you guys, maybe a few of you guys, flew an airplane before. Maybe one of you guys flew a helicopter. Each one of us that operate machines have a license to operate that machine. And if you’re ambitious enough, maybe one of you guys want to do a scooter company or a mobility company. To do that, the people that are operating those machines, they have to have taken a test to operate those machines. To make AI, you don’t have to take a test. There’s no license to deploy systems that are going to reconfigure how people think. And I think that’s wrong. I think that just like operating a vehicle, people that are going to get behind that will should take a test to know if they’re going to just slam up against a crowd of people walking down the street. But we know we’re safe. We’re safe walking down the street knowing that that’s not the first thing that people are going to think of. We trust in humanity. We knew that there were steps taken to ensure that the people that are deploying those cars deployed them with seatbelts, deployed them with airbags, deployed them with sensors. We know that these things are put in place in society. Same is not for AI. We kind of like turn the blind eye that hopefully the adults will be adults. But we know damn well the adults are not being adults. They’re being greedy. And that’s wrong. So even before we get to the Constitution, there should be some basic like, yo, is that person that’s deploying that product, do they got our best interest in mind?

will.i.am: Do they care enough? Or do they lead with greed? Are they human first or money first? Because Web 2, social media, if you look at like the good it did, it did a lot of foul stuff. So AI for good, you also have to do like data rights. What is your data practice? Because you can’t have AI for good with poor data practice. And we are still in process of doing that. poor data practice from, you know, social media. So is this just a slogan?

Nicholas Thompson: Are these just like word logos? This sloganism? Yeah, for good, poor data practice. Guess what you’re gonna have? Fucking bad AI. I didn’t think that that particular group of words would get a round of applause here at the UNAI for good, but I’m delighted it did. When you look at AI as it’s being deployed today and as it’s being developed today, do you think that the people who are doing the most of it are leading with greed, or do you think they’re leading with empathy?

will.i.am: Are they making it for humans, or are they making it for money, as you just said? Dario’s one of the good guys, and I’m happy there’s a Dario. He’s transparent. Demis is one of the good guys. I hope he could get Google to follow his pureness. Mustafa at Microsoft is one of the good guys. So in these mega companies, they have some pretty awesome humans there, two of which were homies growing up, Mustafa and Demis, right? Mustafa’s from freaking, what’s that city in the UK? He took me there a couple of days ago. Anyways, I just brain farted. I thought he’s from London, but he’s from somewhere else in the UK. No, he’s from London, but he’s from a sub city in London that’s like pretty the hood. Aiden from Cohere is the nicest person I’ve ever met in a very, very long time. So there’s a lot of good-hearted folks in AI. And then you have this, the rigmarole that they have to function in, but that rigmarole needs to, they’re either gonna have to force to be more human, but right now it’s kind of like, let’s see how it goes if we lead with greed. So you can’t play with it.

Nicholas Thompson: So you have a lot of influential people in this room. You have regulators, you have government officials, you have builders. If there’s this moment we’re in right now where AI could go in a direction of greed or go in a direction of empathy, what is the role for the different people in this room? How should they think about their responsibilities or what they can do today or tomorrow? Okay, so the folks that are, they’re on the side of innovation, which is great, because you don’t wanna have regulation that stifles innovation.

will.i.am: We wouldn’t be here if we practiced strict, strict, strict regulations 10 years ago. So I understand that. It’s the same with censorship, censorship with lyric writing. You’re on the side of the talent. You know the talent means well. This is how they feel. This is what they see. Let’s make sure we protect them. That’s awesome. And then there is, hey, what about the people though? What about the kids? Should they be around that foul language? What about the kids? Should they be subjected to this camera there? Should every movement be mapped? Should you use their movements against them later on when their machine learned down to blink? Down to the smirk when they scrolled? Down to their search? Like the cookie, damn, they picked the cutest name, cookies, when it’s really a cookie monster. So you wanna make sure, we haven’t been here as a society. This is the most, trying to explain it without sounding spooky. Have we been here before? Or is this the first time we’ve been here? Because it kind of sounds like things that were in fantasy movies in the past. It kind of sounds, if this was like 40 years ago, this kind of sounds like fantasy. You know, there was a word called abracadabra that comes from Aramaic, which means I create as I speak. That kind of sounds like chachi piti. So how does folklore and myth rhyme with what we have now? Like genies and bottles, kind of sounds like, kind of sounds like what we are now. So if we’ve been here before, why did they put the genie in the bottle? And then why are we letting it out? And what are the precautions so that we’re safe now if this rhymes with the past? Right? So I don’t think it’s something to play with. I don’t think it’s something to, but I know that, I know why we’re here. This is a time where we can, we will catapult as a species. We will catapult as a society. We, if we can be humane, but we are the most inhumane thing to humans. Think about what we’ve done. We’ve been able to tame the wildest bears, the wildest lions. We’ve been able to look at an eagle and be like, ooh, I like how they fly. We make an airplane. Ooh, wow, look at those dolphins and them freaking whales. We made submarines. We figured out how to then look at the world, figure out whether, down to the point where we can predict it down to looking inside of us and figuring out our neural network and synthesizing. We are pretty profound. So now what about protocols? Are we that innovative to where we can also put some constraints at the same time as we innovate? I think we could. We put constraints on every wild animal to where we are safe, to live in the mountains next to mountain lions. We are pretty freaking awesome. We are great. Now, where’s our greatness when it comes to leadership and leadership and leadership and leadership? Where’s our greatness when it comes to leadership and precaution and safety? We need that. Because we don’t want the AI to do that. Where else? All right. That is the perfect note to end on.

Nicholas Thompson: I would ask the Swiss government to please book the Black Eyed Peas next year, simultaneous with this concert, so that Will can be here, not in Paris. But thank you very much to one of the great AI philosophers of our time, Will.I.Am. Thank you.

Agreement points

Human-centered approach to AI development and education

Speakers

– will.i.am
– Nicholas Thompson

Arguments

AI should be used as a tool for generating variations of original human ideas rather than replacing human creativity and imagination

The principle of humans generating core ideas while machines elaborate should be maintained across educational fields

Summary

Both speakers agree that humans should maintain control over original creative and intellectual processes, using AI as a tool for elaboration rather than replacement of human cognition and creativity

Topics

Future of work | Online education | Capacity development

Critical moment for AI governance and direction

Speakers

– will.i.am
– Nicholas Thompson

Arguments

AI development should prioritize human welfare over profit, with developers being ‘human first’ rather than ‘money first’

The current moment represents a critical choice between AI development driven by greed versus empathy

Summary

Both speakers recognize that we are at a pivotal moment where AI development could take fundamentally different paths based on whether it prioritizes human welfare or profit

Topics

Human rights principles | Digital business models | Consumer protection

Importance of proper AI education and skills development

Speakers

– will.i.am
– Doreen Bogdan Martin

Arguments

Young people need to develop audaciousness, ambitiousness, and competitiveness as core tools for the future regardless of their field

Will.i.am serves as a special ambassador for the AI Skills Coalition, bringing unique perspective to AI education initiatives

Summary

Both speakers emphasize the critical importance of preparing young people for an AI-driven future through proper education and skills development programs

Topics

Online education | Capacity development | Digital access

Similar viewpoints

Both speakers advocate for maintaining human agency in creative and intellectual processes while leveraging AI as a supportive tool for elaboration and variation generation

Speakers

– will.i.am
– Nicholas Thompson

Arguments

AI should be used as a tool for generating variations of original human ideas rather than replacing human creativity and imagination

The principle of humans generating core ideas while machines elaborate should be maintained across educational fields

Topics

Future of work | Online education | Intellectual property rights

Both speakers acknowledge the complex challenge of balancing innovation with safety and human welfare, recognizing the legitimate concerns of multiple stakeholders

Speakers

– will.i.am
– Nicholas Thompson

Arguments

There’s a tension between protecting innovation and ensuring public safety, similar to debates around content censorship

Regulators and government officials have responsibility to balance innovation protection with human welfare

Topics

Data governance | Human rights principles | Consumer protection

Unexpected consensus

Need for AI licensing and regulation despite innovation concerns

Speakers

– will.i.am
– Nicholas Thompson

Arguments

AI developers should require licenses similar to vehicle operators, given AI’s power to reconfigure human thinking

There’s a tension between protecting innovation and ensuring public safety, similar to debates around content censorship

Explanation

Despite will.i.am being a creative artist who might typically oppose regulation, and Thompson being a tech journalist who often covers innovation, both agree on the need for AI regulation and licensing. This consensus is unexpected given their backgrounds and the typical tension between creative freedom/innovation and regulation

Topics

Data governance | Consumer protection | Human rights principles

Historical and philosophical perspective on AI development

Speakers

– will.i.am
– Nicholas Thompson

Arguments

Current AI developments parallel historical myths and folklore about genies and magical creation through speech

Will.i.am forecasted current AI capabilities in music 15 years ago through his music video, demonstrating early vision of voice cloning and language models

Explanation

Both speakers engage with the deeper philosophical and historical implications of AI, moving beyond technical discussions to consider mythological parallels and long-term foresight. This level of philosophical engagement is unexpected in a tech policy discussion

Topics

Cultural diversity | Interdisciplinary approaches | Human rights principles

Overall assessment

Summary

The speakers demonstrate strong consensus on maintaining human agency in AI development, the need for proper governance and education, and the critical nature of current decision-making moments for AI’s future direction

Consensus level

High level of consensus with significant implications for AI policy – all speakers agree on human-centered approaches, the importance of education, and the need for responsible development practices. This consensus suggests a clear path forward for AI governance that balances innovation with human welfare, emphasizing the urgency of implementing proper safeguards and educational initiatives.

Different viewpoints

Unexpected differences

Overall assessment

Summary

The discussion shows remarkably high alignment among speakers with no direct disagreements identified. The speakers share common concerns about AI governance, human-centered development, and the need for proper regulation.

Disagreement level

Very low disagreement level. The conversation demonstrates consensus on major issues including the need for AI regulation, maintaining human creativity, prioritizing human welfare over profit, and the importance of proper education and preparation for AI futures. This high level of agreement suggests strong foundational consensus on AI governance principles, though it may also indicate limited diversity of perspectives in this particular discussion format.

Partial agreements

Similar viewpoints

Both speakers advocate for maintaining human agency in creative and intellectual processes while leveraging AI as a supportive tool for elaboration and variation generation

Speakers

– will.i.am
– Nicholas Thompson

Arguments

AI should be used as a tool for generating variations of original human ideas rather than replacing human creativity and imagination

The principle of humans generating core ideas while machines elaborate should be maintained across educational fields

Topics

Future of work | Online education | Intellectual property rights

Both speakers acknowledge the complex challenge of balancing innovation with safety and human welfare, recognizing the legitimate concerns of multiple stakeholders

Speakers

– will.i.am
– Nicholas Thompson

Arguments

There’s a tension between protecting innovation and ensuring public safety, similar to debates around content censorship

Regulators and government officials have responsibility to balance innovation protection with human welfare

Topics

Data governance | Human rights principles | Consumer protection

Key takeaways

AI should augment human creativity rather than replace it – humans should generate original ideas while AI helps with variations and elaboration

Future-ready education requires developing audaciousness, ambitiousness, and competitiveness as core skills regardless of field

AI governance needs fundamental reforms including licensing requirements for AI developers and an AI constitution with clear principles

Current AI development faces a critical choice between being driven by greed versus empathy and human welfare

Good data practices are essential for beneficial AI, as poor data practices from social media continue to cause problems

While individual AI leaders may be well-intentioned, they operate within systems that currently prioritize profit over human welfare

Humanity has successfully managed dangerous forces before and can develop appropriate constraints for AI while maintaining innovation

Resolutions and action items

Will.i.am appointed as special ambassador for the AI Skills Coalition

Implicit call for developing an AI constitution that organizations must sign to demonstrate trustworthiness

Need to establish licensing requirements for AI developers similar to vehicle operators

Requirement to evaluate AI developers based on whether they prioritize human welfare over profit

Unresolved issues

How to balance innovation protection with public safety in AI regulation

What specific principles should be included in an AI constitution

How to implement licensing requirements for AI developers in practice

How to ensure good data practices across the AI industry

How to shift AI development culture from greed-driven to empathy-driven

What specific constraints should be placed on AI development while maintaining innovation

How to address the continuation of poor data practices from the social media era into AI

Suggested compromises

Regulation should avoid stifling innovation while still protecting public welfare, similar to content censorship debates that balance artistic expression with public safety

AI can be used as a tool for generating variations of human ideas rather than completely replacing human creativity

Recognition that both innovation advocates and public safety advocates have valid concerns that need to be balanced

AI now has this metaphor. AI now does Pilates. AI now stretches. That doesn’t mean you don’t have to stretch your damn self. You still got to stretch. I don’t care how dope AI is at stretching and doing Pilates. You still have to be flexible yourself.

Speaker

will.i.am

Reason

This metaphor brilliantly captures the fundamental relationship between human creativity and AI assistance. It challenges the notion that AI tools can replace human development and skill-building, emphasizing that personal growth and creative stretching remain essential even as AI becomes more capable.

Impact

This comment established the philosophical foundation for the entire discussion about human-AI collaboration. It shifted the conversation from technical capabilities to the deeper question of maintaining human agency and creativity, leading Thompson to explore how this principle applies to education and creative processes.

We are the ideators. It is an infinite generator. At this point in time, it’s just regurgitating imagination. It’s a regurgitation of imagination. We are the dreamers of the dream, Willy Wonka.

Speaker

will.i.am

Reason

This comment provides a profound distinction between human creativity and AI capability, positioning humans as the source of original ideas while AI serves as a tool for elaboration. The reference to Willy Wonka adds cultural depth, connecting the discussion to broader themes about imagination and creation.

Impact

This insight became a pivotal moment that led Thompson to explore educational implications and ask whether this principle should guide how we teach students to interact with AI. It elevated the discussion from practical AI use to fundamental questions about preserving human cognitive abilities.

Each one of us that operate machines have a license to operate that machine… To make AI, you don’t have to take a test. There’s no license to deploy systems that are going to reconfigure how people think.

Speaker

will.i.am

Reason

This analogy powerfully illustrates the regulatory gap in AI development by comparing it to well-established licensing systems for operating vehicles. It highlights the paradox that we require licenses for physical machines but not for systems that can reshape human cognition and society.

Impact

This comment dramatically shifted the conversation from creative applications to governance and regulation. It provided a concrete, relatable framework for understanding AI regulation needs, leading to deeper discussion about greed versus empathy in AI development and the responsibilities of different stakeholders.

You can’t have AI for good with poor data practice. And we are still in process of doing that poor data practice from social media. So is this just a slogan? Are these just like word logos? This sloganism?

Speaker

will.i.am

Reason

This comment cuts through corporate rhetoric to expose the fundamental contradiction in promoting ‘AI for good’ while maintaining exploitative data practices from the social media era. It challenges the audience to confront whether their commitments are genuine or merely marketing.

Impact

This created a moment of uncomfortable truth-telling that energized the audience (noted by Thompson’s surprise at the applause). It forced the conversation to confront the gap between stated intentions and actual practices, leading to more honest discussion about whether AI leaders are motivated by greed or empathy.

Have we been here before? Or is this the first time we’ve been here? Because it kind of sounds like things that were in fantasy movies in the past… abracadabra that comes from Aramaic, which means I create as I speak. That kind of sounds like chachi piti… genies and bottles, kind of sounds like what we are now.

Speaker

will.i.am

Reason

This comment provides a fascinating historical and mythological perspective on AI development, suggesting that current AI capabilities echo ancient human dreams and warnings. By connecting AI to folklore about genies and creation through speech, it adds profound depth to understanding both the promise and dangers of AI.

Impact

This shifted the entire discussion to a more philosophical and cautionary tone, moving beyond technical and regulatory concerns to existential questions about human nature and historical patterns. It provided a powerful framework for understanding why AI requires careful handling, leading to the discussion’s climactic call for leadership and precaution.

Overall assessment

These key comments transformed what could have been a standard tech industry discussion into a profound exploration of human creativity, responsibility, and wisdom in the AI age. will.i.am’s insights consistently elevated the conversation from practical concerns to fundamental questions about human nature, governance, and our species’ relationship with powerful technology. His unique perspective as both a creative artist and tech entrepreneur allowed him to bridge abstract concepts with concrete examples, while his willingness to challenge industry rhetoric created moments of genuine insight. The discussion evolved from creative applications to educational philosophy to regulatory frameworks to existential questions about human civilization, with each key comment serving as a stepping stone to deeper understanding. The overall effect was a conversation that honored both the transformative potential of AI and the essential need for human wisdom in guiding its development.

How should different stakeholders (regulators, government officials, builders) think about their responsibilities in ensuring AI develops in a direction of empathy rather than greed?

Speaker

Nicholas Thompson

Explanation

This question was posed to will.i.am regarding the roles and responsibilities of influential people in the room to guide AI development toward human-centered outcomes rather than profit-driven ones.

What are the specific principles that should be included in an AI constitution?

Speaker

Nicholas Thompson

Explanation

While will.i.am mentioned the need for an AI constitution, the specific principles and framework for such a constitution were not fully detailed and require further exploration.

What would an AI licensing system look like in practice?

Speaker

will.i.am

Explanation

will.i.am suggested that people deploying AI systems should need licenses similar to driver’s licenses, but the practical implementation of such a system was not explored in detail.

How can society implement constraints and safety measures while maintaining innovation in AI development?

Speaker

will.i.am

Explanation

will.i.am posed this as a challenge for society – whether we can be innovative enough to put constraints and safety measures in place simultaneously with AI innovation, similar to how we’ve managed other powerful technologies.

What are the historical parallels between current AI development and past folklore/myths about genies and magical creation?

Speaker

will.i.am

Explanation

will.i.am raised intriguing connections between AI capabilities and historical myths (genies in bottles, ‘abracadabra’ meaning ‘I create as I speak’), suggesting this area deserves deeper investigation for insights about managing powerful technologies.

How can ‘AI for good’ initiatives address poor data practices inherited from social media?

Speaker

will.i.am

Explanation

will.i.am highlighted that AI for good cannot succeed with poor data practices from the Web 2.0/social media era, but specific solutions were not detailed.

How should the principle of human-led ideation versus AI elaboration be applied across different educational fields beyond music?

Speaker

Nicholas Thompson

Explanation

Thompson asked whether will.i.am’s principle of keeping core creative ideation human while using AI for elaboration should be applied universally in education, but this broader application was not fully explored.