will.i.am and Matthew McConaughey on Keeping AI Human | Dreamforce 2026

Thank you everyone for coming. I'm Suzanne DiBianca, Chief Impact Officer at Salesforce. We are really excited to have this conversation today. We— I looked at the title over the weekend and I thought, you know, should we change this title from AI and Humanity to AI or Humanity? We realized let's just keep it as is, but we wanna touch on a little bit the news of the weekend. So my fellow panelists and friends are amazing humans. So thrilled, been here. How many, what Dreamforce is this for you?

How many Dreamforce? Oh, somebody help me out. Camilla, Levi? Quite a few. Anybody? Mark? Quite a few.

My guess is 5.

5-ish. 5-ish, yes.

This is my 12th.

Okay. All right. You're up there. Yeah.

We're 2026?

Yeah. Yeah.

That's the year.

Yeah, great. So I think we should just jump in. Let's talk about what happened over the weekend as it relates to AI and humanity, as it relates to the tweet that sort of went out with this 10% number. and the essay that was published by Dario around Pacing the Frontier. And I'm curious just sort of on your take on that news and what you think about it.

A couple of things. First off, you know, I think when we first started to engage with AI a few years ago, there was always this— the word I always heard was it has no agency until you prompt it. And while it's still dumber today than it will ever be, It's self-prompting. I think going forward, there's this word that we keep saying about regulation, and just a minor prescriptive tweak in the word, but I— and some people are using it, but I think it's very smart if we start using the word standard instead of regulation. Obviously, regulation feels like a mandate. No one likes to be told what to do, but a standard is something you're embarrassed not to do. And I think a lot more people will be more conscious of how to make the standards. and to live up to them if we use that simple term as— it's almost a marketing tool. But yeah, obviously it's time. What will change? When will the standards be? I think a disaster, some sort of disaster will have to happen first. And that accident is what will force us to create certain standards. I hope it doesn't happen that way, but I believe it will.

And standards, you mean consistency? You mean thinking about this idea of this third-party auditing? Some guardrails. Talk about what you mean by standards in this context.

Well, so right now, mainly AI is created to produce more faster for financials. Fine. But that's not necessarily a standard that's going to improve civilization and life. There needs to be, I believe, some values and ethics and guardrails, as you say, that should be in there for the— some culture needs to be put into it and understood. And again, I don't think we'll actually do it until a real accident happens. But again, we're getting wake-up calls like this last couple of weeks. Now, who's supposed to do that? Who's supposed to put those values in there?

Yeah.

I'm not sure, but somebody or some people who are accountable will be accountable and hold themselves accountable for expanding the definition, our definition of profit. A profit cannot only be numerical and cannot only be quantifiable. It has to be qualifiable. And we have to— that's what I mean by the standards and ethics and values that need to come in there, is how can AI amplify the life force? Needs to be something and not just the number, because if it's just the number, we see what's going to happen. A lot of people are going to be left behind and problems are going to happen. Big problems.

How about you, Will?

So I'll go a little bit more simpler as far as things that are put in place. Let's say, for example, there were car companies and not AI companies, and they're deploying vehicles that go 0 to 60 under 1 second, and there's no seat belts, there's no airbags, And would that car be allowed on the streets? No, right?

Yep.

Would you be able— you could buy it, but can you drive it? Will you get pulled over going like 0 to 60 in under 1 second and sustain that speed? And now you're in a school area, will you get pulled over? You will, right? OK, cool. Can I speed just through the crosswalk at a red light? No. So in our cities, in our world, there are these regulations that car companies adhere to. And I think it's just a matter of societal, community, human decency to have some basic things put in place.

Mm-hmm.

to ensure that we're safe. And I love tech, it's freaking awesome, it's amazing, but not at the cost of like, to stable— not having a stable society and just waiting for when something goes wrong and then we're gonna have regulations.

Yeah.

That's irresponsible, you know? And for example, if these companies came outta Watts And the folks that built these awesome technologies just grabbed up all the data in the world. That probably would not be the case if the folks that built it came from Watts.

Mm-hmm.

So you just got to keep it real, keep it 100, and make sure people are safe while we enjoy and see the world change with these technologies.

Is it fair to say we have— there's about a dozen men, industrialists, that our future is in the eyes and the hands of right now that were unelected?

Yeah.

Um, and I'm not saying that as a, as a, as a, as a warning or a downer. I'm just saying that's, that's the facts, isn't it?

That's right. And this is a technology that needs to be in the hands of all humans. And so I think what I'm really hearing is Sort of agree. Like, let's pace this to get it right. There could be some crisis that comes up that will force us to do it. But if we have the opportunity now to put the seatbelt in the car, it's a good idea to do it before somebody gets hurt.

Or airbags.

Or airbags. You just said— And/or.

Or auto-lock brakes.

This is technology for everyone, though. It's not going to be equally distributed though, because the power, electricity will cost too much to equally distribute it, correct? Back to your watts analogy, right? Will not be equally distributed democratically. I mean, am I correct in saying that? I believe I am.

I think so.

I mean, it's more of a power question for how— that's also one of the questions. I mean, and how it's going to be across and accessed and scaled across society.

Yeah.

But it is gonna leave some people behind.

And that's one thing it, it could, it has the potential to. And when we started this work at Salesforce, as we were thinking about education, it was sort of in the age of the digital divide, right? Where some people had computers, some people had internet access, some people didn't. And so I think we have a real opportunity to not go in that direction today. And Will, you've always been an incredible, advocate for access. And for those that don't know, Will has been teaching a class at ASU. It's your second semester you're going into?

Yep.

Called The Agentic Self.

The Agentic Self.

The Agentic Self. So as you think about both access and builders, right, you're a builder, you're a tinkerer, you've always been. How do you think about this idea of access To make sure we don't end up where we were before in the digital divide? And like, what did you learn from your students?

My hunch was, hey, what if we brought this technology and put it in the hands of students? And borrowing from the paper that got us here, Attention Is All You Need, the transformer paper, What if we tweak that title with intentions is all we need for transcendence and encourage students to run at a problem with an agent they build so that they can solve that problem away? And that was our gut instinct for Unit 1. And we announced the findings from that at ASU GSV in April. We started in January, right before we had you know the last class, we presented our findings at ASU GSV, which is like Arizona State University's big tech summit, like the EdTech of Dreamforce, the Dreamforce of EdTech. And there I sat on a stage like this with Chancellor Sonia, who is the who is the chancellor of California Community Colleges in the audience. And she pledged to take the Agentic Self course and bring it to California Community Colleges alongside our edu.fyi agent builder tool. So at FYI, we created this very simple configurable agent builder for students, faculty, and students in collaboration with ASU and their EdPlus unit there, and we called it Edify. So now what we're doing with California community colleges alongside ASU is we have this book of problems. Then just like the title, it lists problems that are affecting inner city so that students, residents from those communities solve those problems themselves. So as far as access, Compton College, folks that go to Compton College are from Compton. Folks that go to Pasadena Community College are from Pasadena. Very rarely does somebody that lives in Brentwood go to freaking Compton, right? I went to Santa Monica Community College only because I went to Pacific Palisades, and that is like the trajectory from Pacific Palisades, Santa Monica College, and if you go there, you can go to UCLA. My point is we want to go at community colleges because in these communities are gold mines.

Hmm.

What do I mean by that? I mean, just like Uber rendered a lot of inner-city conundrums obsolete, you gotta think of what Martin Luther King and the freedom fighters that fought for civil rights were able to accomplish. And thank God they did. Thank you for your service. Thank you for your contributions, all the freedom fighters and the folks that fought for civil rights. But no matter how hard the fight, Taxi drivers didn't stop for people of color in the '70s, '80s, and '90s, 2000s. And then the iPhone, GPS, and Uber solved those problems. Now people don't complain about, man, they didn't pick me up like they did in the '90s or in movies. You would see it in movies and TV shows, taxi, and they wouldn't stop for people that look like myself. Now that's not the issue. Now, how many other problems are in our inner cities that can be solved and rendered obsolete when you look at access as like, no, they're underserved communities, underdeveloped communities, and this time we are the developers.

Mm-hmm.

The folks in Compton are the developers. They are the solution bringers. They are the servers. And the agent that they build at Compton Community College is gonna be put to work. And that is what we mean by, like, equal opportunity, by giving folks the dignity and mentoring, encouraging them to go out there and be intentional and solve problems in the community that they come from. We are going to redefine the whole concept of what community college is in the Angentic State. And so thank you, Chancellor Sonja, for your leadership and partnership.

That's cool, man. Outstanding. Piggyback on that. You talk about intentions for solving the problems. Right on. Because, and again, when I say what I'm about to say, this is not me being an AI pessimist. It's just something that I've heard that I want to share that's going to highlight why you need intentional intentions like this. I've interviewed some heads of some of these major tech companies, and when I asked, okay, you're creating the new in a biblical sense, the new Eden. What are the rules? What are the rules of engagement? What's the expectations? What's the culture? What's the ethics? The answer was, the algorithm will tell us.

Hmm.

And that scared me. And it's the antithesis, I believe, of what you're saying with creating in these communities people that are intentionally building to solve problems with AI, which is very different than saying, I'm stepping out and taking the Fifth on this one, which is good for us to just know that we are going to have to take control of it, either personally or in our own communities, to make it work for us.

Yeah, that's right. And I think, and like you, I'm a huge fan of the community college system. Feels to me like one of the most underrated resources in our country. And I'm really glad to hear that that's one place that you're focused as it relates to access. And one of the things that I learned to be sort of an AI technology optimist is I think AI has a lot of potential in the nonprofit sector. to be able to scale. So it might be in the hands right now, the building of it in terms of, you know, 10, say, companies, 3 mega ones, right? But if we can do our work well, we can enable these small nonprofits that are resource-strapped to be able to do tutoring and mentoring at 11 o'clock at night once the volunteers have gone home. So I see a lot of potential. But it needs intentionality to get there. And I was thinking, Matthew, I heard you talk about AI, thinking about it as like a mirror.

Yep.

Right? And so I wonder if you can just talk about what that means to you, like why you use that analogy and what it might mean in this context.

On 3 levels. Me personally, I'm using it to try and get to know myself better. So I want a true mirror of myself so I can remember stuff I've forgotten. I can compound and scale and associate my own thoughts and theories and find out where they worked in history and where they can work in the future, and they feel like they're making me smarter. Secondly, Instagram is its own mirror, but it's one of those thin mirrors like the ones that are in the retail shops in Beverly Hills or the ones that are in our favorite gym that make us look a little thinner, And a little better, so we'll buy the dress and keep going back to the gym. We love those mirrors, but they're not a true mirror. They're overly optimistic mirror.

Yes.

Right? AI has given us a chance to create— we're going to create an absolute true mirror of ourselves, and we may not like what we see, but it will be an absolute— everything we ask it and everything in ways that we do or do not challenge what it prompts back to us, the answers it gives are going to give us consequences that we are going to be looking ourselves square in the eye like never before.

Yeah.

And that's very different than a social media IG or— but we are going to be looking at ourselves. And that's just a wonderful responsibility to look at and a privilege that we have right now in creating it and thinking about, which I keep going back to, think of the answers are free and cheap, man. The questions are not. The questions are scarce. So what questions we ask it and to also challenge what it spits back to us, which seems so certain.

Yeah.

Sometimes you're like, don't touch that. Did I say that? Did that help me say that on my own? It did. I want to put that in quotes and engrave it and put it over my bed. No, hang on a second. Challenge it because it's just giving you patterns of trillions of things that it's getting. So it's not personal yet. So challenge it. Think about the questions you're going to ask, and then make sure you challenge what answers it gives you.

And one of the things that I did with my with the AI that I use is I put in sort of a mega prompt, which is, don't always agree with me. Challenge my thinking. And if you don't know the answer, say you don't know the answer. And it's actually really shifted the results that I get. But these are sort of 3 simple things. Like, I love those mirrors. I do too. Right? But they're not a true reflection of yourself. And I think you can't learn and grow unless you get sort of a real reflection. And like you said, you might not always like what you see, but in this context, you're going to get the rosy story unless you say, don't give me the rosy story all the time.

Yeah.

And like, I wonder with you, Will, in your experience in the classroom and outside of the classroom, I've seen your tech, I've seen some of what your students have built. Has that happened with you where you get Sort of an overly optimistic agent that's not really grounded in reality? And then what do you do about it?

So the product that we have at FYI.ai and in collaboration with ASU, edu.fyi, is that our product, when you build your agent, one, we don't— want to onboard users. We want to onboard owners. So that agent that you build is yours. The data is yours. And so the configuration of what it believes, what's its concerns, what its aspirations are, its goals, its mission, its point of view, its perspective. Configuring it like that is super important. And the mission, where has it taken you? For example, when you type in GPS, your current location to LAX, it spits back 3 options. In reality, it gave you— there's a million back there.

Sure.

but it gives you 3. The same is for like these AIs. There's all these different potentialities. And if the business model, if companies' business models in the past were never pro-human, you gotta be careful.

Mm-hmm.

With giving yourself Like literally giving yourself to a company. It's like digital virtual prostitution at its highest order because it's the most intimate side. You're sharing how you think. You're sharing your very intimate— it could predict you.

Yeah.

And if If you're going to give something to a company and your predictions, you better look into the terms and conditions and have your lawyers actually check it out because none of the shit in the past did your attorneys read any of those fucking contracts. And if they did, you wouldn't be on those fucking platforms. So we cannot make the same mistakes from social media to agentic media.

Mm-hmm.

You can't go from like, you know, I hope not. I hope this next path that we're in is not a repeat rinse of social media and the agentic state. I hope the business models are not the same. I hope the freaking, you know, civil liberties and data practice and selling your precious data to like, foreign countries that just so happen to not like us. I hope that ain't the fucking tomorrow.

Ditto.

So it is your agentic self. Hence the course. And what we teach is like ethics at the core, self-governance at the core, because, you know, we're accountable for— and shifting that accountability to the folks that are going to be building these tools, holding the student responsible and encouraging them to do the right thing. That's why we provide the book of problems. Either— what you don't want to have is companies and AIs solving humans' problems without the human.

Right.

That is horrible. 2046 looks like this. And the AI said, the problem for global issues is us. Okay? Why would we want that? Why would we want some company's AI solution to point out the reason why our communities are in shambles is because of us? The environment is in shambles is because of us. No, no, no, no. That would be the worst nightmare. And that's what the movies are about. The movies of AI taking over is because it found out we are the reason. We're the reason for being inhumane to humans, freaking plaguing and fucking destroying our own environment and the planet that we live on. How fucking low.

Right.

So we need to be more humane. We need to be a little bit more empathetic, a little bit more— have more tolerance for differences. collaborate better, and allow this opportunity to bring out the best human age that we've ever fucking seen.

Agreed.

So, so, Will, are you saying that in 2046 it is going to say this happened because of us? You're saying now it's up to us to let alongside of the things that we will inevitably screw up, Also point out a whole lot of things where we improved and evolved and progressed. No, because we will be looking in that mirror.

2046, somebody has a kid just now, they're 1, and they're going to be 20 years old. And that new 20-year-old, 30-year-old leaders, AI is going to be like, Oh, wow. You humans are really amazing. Look at all that power efficiency. You guys eat plants, utilize your brain, your neural network. Look at the power efficiency of you guys, humans. You guys created us.

Mm-hm.

Look at how you guys just— took this opportunity, we predicted that you guys were gonna fuck up. Our predictions told us that you guys are not gonna get out of your way. You guys are gonna continue to extract from the planet, use it to hurt one another, just shoot each other, be mean to each other on social media, say the foulest stuff, the algorithm dividing you. We thought we could continue to divide you guys, but this next generation proved us wrong. You guys are the most angelic, kindest, thoughtful, spiritual beings that we can never be.

Hmm.

Here we are having to, you know, we're plastic. Even though we try to like do organic versions because technology would've advanced further from there, AlphaFold, would be like printing out organic brains. But us natural ones, we're going to be all right. And isn't that the future that we want to be running towards? Or are we just going to be running to like swipe, swipe, tap, tap, swipe, swipe, bye-bye, swipe, swipe, hate, hate, fuck you, fuck you, bomb, bomb, bomb? What the fuck is that? That can't be fucking tomorrow.

Yeah.

That's horrible. Wait, stock market, stock market.

Uh uh uh uh.

What the fuck? We motherfuckers ain't got no jobs.

Right.

How they fucking gonna buy shit?

Yeah.

What are they gonna watch when the China model already came out? That's like I guarantee you in 2028 there's gonna be new China Kimmy point four point zero. It can make movies and it just made the whole fucking new Batman series. The whole fucking shit, fuck outta here. I could see the one that it's already making music better than, well, not better than my shit, but yeah, almost.

Yeah.

It's fucking, it can't be. So I think what's gonna happen is like stage. Yo, I went to go see them live. I went to that theater. That guy was improving. Oh my gosh, he was so awesome. It's gonna get back to like physical this.

Well, that's already happening too. Live events, sports.

Yes.

And music. People are flocking to that. And that's not just because we're coming outta COVID.

So what do you think, you know, you write and think a lot about purpose and meaning. What do you think in your sort of analogy in 2046, what do you think is uniquely human? that AI will never be able to do? And where should we be thinking about doubling down as humanity?

Imagination. Imagination, like this building. Somebody imagine this. This building was made, this couch, this rug didn't go through some algorithm. There's a couple of things that are still sacred. And the physical world has not gone through a machine except for these buildings that withstand earthquakes that were done in a simulation, the iPhone, all— what is the point of a simulation actually? Have you ever seen machines before? To make a product, you put it through a simulation, you do all these drop tests, these break tests, You drop it a million different times and out comes an awesome product. So 2046, the one thing that machines can't simulate— unfortunately, it's going to simulate a lot of awesome things. But the one thing is like the air, the beauty of touch, And space and smell and laughter and like, what did you just say? Say that again. And seeing somebody's face light up and these micro moments of like, ha ha ha. And just falling in love with the moment and then leaving and drunk off of that moment. And just the— because we're all just swiping through fucking phones. No one's paying attention to everything. Our mind has been reconfigured to be like notifications. No one's paying fucking attention to anything. And it's gonna have to get back to like every moment is special and precious. So when you see your aunt, because what I'm worried about, the way I looked up to my grandma, like, oh, Nanny, she knew everything. She was born in 1926 and then she passed away. And now when I'm like 80, are the youth, are they gonna look up to like 80-year-olds?

Mm-hmm.

Because they're gonna have some like all-knowing thing in their pocket. Are we gonna look— are they gonna look up to the adults, the elders, like we used to look up to our grandparents? At some point in time, we need to all start teaching what's sacred. What do we value? We only value stocks, stocks, market, money, money, money, money, money.

Stuff.

The machine was gonna learn from that and be like, well, you guys don't value you guys. Come on, you're rich. The machine is going to learn that we don't value these little micro moments. And just like getting old and wrinkles, what happens to the gums and like, say that again, uncle. Damn, I love that your breath stinks. All these little things. Because every day you watch it. After here, you're going to leave here and just start looking. and absorbing and be critical on how we behave and how you behave and how much of the in the moment are you.

Hmm.

And how much has the notifications conditioned on how you are active in that moment, how you're contributing to that conversation. We need, we really need to teach our youth that because just giving them the iPad The iPad's now watching. The devices are watching.

And listening.

And learning.

Yep.

And no other animal's doing that. We don't have to worry about the cats like, yo, them fucking cats is watching, bro. Fucking birds is watching. The machines are watching. That's— the machines are watching.

So imagination. Micromoments, staying sort of present with each other. Matthew, you said people are flocking now to sports events, live music events. What is your take on what is uniquely human and what we have to preserve?

That we remain the author and maintain authorship. I mean, we are what we cause to happen. And causing things to happen is not easy a lot of times. And if we let AI do things for us, we'll remove the resistance to get things done. We'll remove an experience and a knowledge and a wisdom we get from getting things done, of failing as well. So I don't want AI to create a life that we're merely spectators in and looking for it to approve us. Which is a version of objectifying our own life, which is not the way forward, I don't believe. Second, questions. I mean, the winners in this game are going to be the ones that ask the best questions, I believe. The answers are cheap, answers are free.

Hmm.

Questions are sacred. It's where wisdom is. Intelligence is in the answers. Wisdom is in the questions that we ask it. And to piggyback on the questions, not only is it the question we ask, but Not to forget to challenge the answer that AI gave us. It's obviously going to be here as an integral part of our life, but to challenge its, its answers is to make them personal for us. If not, we'll just be a bunch of sheep running across, imitating, mimicking the patterns of a trillion people.

Curiosity is in the questions, and I think that is uniquely Human. The time has gone really fast, and I could go on with y'all for another hour on this topic. But as we, you know, kind of close it down, Matthew, your kids are here, my kids are here, we have young people in the audience. What is your advice around skills that they're going to need walking into the future? And how they should relate to this technology before it sort of takes hold?

So kiddos, I would say go— we talked to our children about this, and one of them who's on, like, just got on socials is like, the problem that I have as a parent with socials is I see too many young people waking up in the morning and deciding who they are and what they want to do based on what could get the biggest likes and most amount of response. So they're gauging their day again, they're living off of looking at themselves in the Jumbotron of life to say, oh, what should I do? Rather than doing something and then going, hey, this is really cool. What if we recorded it and shared it? They're still the author that way. One of the underbellies of what we've been talking about here is that the next generation, please study and know the humanities. It's what is going to prompt better questions. It's what's going to make that mirror that we create one that's true and also one that we don't have to just look at and drop to our knees and cry about, that we can also look at and go, there we go. Look what we did. Look what we're doing. See if you can ask questions that prompt answers that inspire you personally. Again, challenge AI. Don't forget, it needs to be pushed. It needs to be pushed. I hope that it expands minds, which I believe it will. And then the caution would be not, again, back to authorship, not owning your own life, you know, not letting AI make you a spectator who you're looking to just to approve. your own life. And lastly, the, the— say it again— go, kiddos, go, go, go live your own hero's journey before you go to AI and let it tell you what your journey ought to be and what's going to happen before you even find out.

Thank you. Powerful. And I'll, I'll give you the final word on that question. What is your What's your advice to young people today as we walk into this world that we're in?

Remember your cloth, who we are as a species. We have been the apex brain on the— alpha brain on the planet. Bears are stronger than us. We tamed them. Elephants are stronger than us. We tamed them. We looked at a bee and was like, wow, look at that. And we made— we looked at a dragonfly and we made helicopters. We looked at whales and made submarines. Humans are awesome. And don't let— don't ever let a company tell you that their product is smarter than all humanity. Fuck outta here.

Thank you.

Well, thank you. Thank you both for your time, for your wisdom. It's a pleasure to work with you. It's an interesting world we're walking into. Yeah, it's exciting. So keep bringing your imagination, keep bringing your curiosity, and We're gonna be okay.

All right. Thank y'all.