(announcer) Here to make the case for AI.

Please welcome Reid Hoffman co-founder of LinkedIn and Manas AI with Atlantic staff writer Josh Tyrangiel.

[insturmental music] [applause] (Josh Tyrangiel) Yes, I apologize.

I'm back.

Reid, it's great to have you here.

(Reid Hoffman) It's great to be here.

(Josh Tyrangiel) So there are many interesting things about Reid Hoffman, but one that I envy is that you're in the sort of position to divide your time and attention into a variety of pursuits, and you've really focused on AI, in a general sense, for the last, I don't know, eight years?

(Reid Hoffman) Maybe even ten.

(Josh Tyrangiel) Okay.

(Reid Hoffman) And my undergraduate majo was an AI, too, but, you know.

(Josh Tyrangiel) Yeah.

Mine too.

Mine, too.

But is it fair to say, that you spent a lot of 2025 focusing in a way that you haven't before?

(Reid Hoffman) Yes.

(Josh Tyrangiel) And where have you focused?

(Reid Hoffman) So, one of the things is, you know, I, with actually, local New Yorker Siddhartha Mukherjee.

You know, the awesome author, celebrated oncology researcher, co-founded a company called Manas AI which is based here in New York.

It's actually one of the things that, you know, I will spend a couple of days working at because I'm a co-founder of this.

And the basic idea is to create a, to use AI to accelerate drug discovery with a particular focus on cancer.

But as you're creating a drug discovery factory, it also allows you to get diseases that are, you know, significant, but not as, you know, kind of the economic model doesn't work for them because there are fewer theyre orphaned, etc.. And the the idea that we, you know, kind of, came, brought, brought about for this was that neither the purely, typical biological pharma approach will work with a little sprinkling AI, nor will the only AI work that you have to reinvent it kind of fusing the two together, because, there's some places where AI is a huge accelerant and helps you create that factory of understanding targets possible.

You know, bindings for those targets, ways of understanding how to, you know, test them before you start getting into clinical testing as a way to make that work.

And, the AI ha to be kind of in that process, but also it isn't like just purely simulated.

And so, it was really funny because th way the company came about is, I had this idea about what are the kinds of things that I could accelerate that most people, in Silicon Valley, aren't paying attention to.

And, drug discovery was one of them.

And so I called Sid, because we've known each other for a number of time.

And I said, hey, come to New York, I'd like to have dinner with you.

He said, “great.” And we had dinner like, so I was I was running through all this stuff, and he said, well, you know, I brought cancer drugs to market, right?

I'm like, no, I didn't know that.

Right.

So, so then that's, that' and then it becomes a company.

(Josh Tyrangiel) So, one thing that I think we were just talking about quantum and, there's a high degree of pretending when it comes to quantum.

I do think there's also a high degree... (Reid Hoffman) There's a super position of pretending.

(Josh Tyrangiel) Yes, yes.

Because there's a degre to which we've all accepted that AI and health care, sort of like the Reese's Peanut Butter Cup of hard problems and advanced technology.

Can you actually just take a minute and explain why they pair pretty well?

(Reid Hoffman) Yeah.

Yeah.

So by the way, there's, medical as a whole range and we can go to any of these things, but there's everything from one of the things I think we want as a human as a human race, as a society, as urgently as possible, as a 24/7 medical assistant running on every smartphone that is already, you know, like, like it's it's, it's essentially doable.

Like if, for example if you get a serious diagnosis and you don't consult ChatGPT or a frontier model as a second opinion, you're making a mistake.

I actually know people's lives who've been saved by that consultation, not first opinion, right?

Second opinion but still extremely important.

So that's the kind of very front end.

There's also, of course, like for example if you're doing medical practice and you're not actually, i fact, uploading your case notes and doing a cross-check second opinion again like, like I, some year that's a that's malpractice.

Just because what are these things?

They have ingested a trillion words and has said I have all this knowledge.

We don't have that.

I mean that's part of, you know, that's part of the amplification we get.

And then you get into, you know, a whole bunch of other like, you know, kind of how do you, you know, kind of operate hospitals, medical care.

You can imagine, for example, with the medical assistant, it goes, oh, you get to the E.R.

right now.

You can imagine scheduling and things like, you know, single payer systems like NHS of oh, as opposed to just, well, you set an appointment and shows up in eight weeks.

It's the hey, you talk to the assistant.

The assistant actually, in fact, triages you in, so you get all that to here then you get to this stuff that, you know, as a tech entrepreneur and, and investor is, part of the thing that you can see what's happening with AI is language and biology, actually in fact, closely approximates.

It's not exactly a language, but there's a lot of language in computation.

And so the notion, for example, like the that the typical cancer therapeutic chemo is we don't know how to do something specific with the one of a thousand cancers you have.

So what we're going to do is poison you and we're going to hope that we kill the cancer before we kill you.

And we have this little toxicology window that we're like, oh, let's make sure we don't kill you, but let's try to kill it.

And that's what we're trying to do.

Well, that's because we we haven't, crafted the specific molecule that would say your particular kind of cancer.

How do we, you know, figure out a molecule that will bind to that, only bind to that, so not kill you.

Kill the cancer cells, but not kill you.

And and then that that becomes a huge search space problem in language.

And so how do you use AI, which is search based on language to accelerate the understanding of what are the things that would bind that?

Because, by the way, it's not part of the binding question for the folks who, you know, may not, I've now learne a whole bunch of drug biology, it's not just bindin with the specific cell itself, but also what else does it bind with?

What else might it kill?

And you have to make sure you don't have those negative effects.

(Josh Tyrangiel So you as a serial entrepreneur you sort of launched companies, you've entered different industries.

Tell me a little bit about wha it's like to enter health care.

And how is your approach different from Big pharma and from interests that have been there for a really long tim and are super well capitalized?

(Reid Hoffman) Well, so, I woul say that my normal thing as a, you know, kind of Silicon Valley entrepreneur is avoid the regulated industries because anytime you have this kind of regulation that slows down innovation, massively increase complexity, makes it more likely you're going to fail as a startup, you know, harder to get financing a whole bunch of other things.

So part of my, kind of I wouldn't have ventured into this except with Sid, because Sid knows all that part of the world extremely well.

But, it's oh, I have some differential knowledge about AI that could make a difference here.

And this makes a difference at a, five year olds die from cancer.

I mean, it's like it's it makes a difference at a humanity scale.

So I will navigate all of what looks to me like overly ornate craziness around regulatory industry, how economics flow, how clinical trials work, you know, how all this stuff works.

But we'll we will get through that because the outcome is potentially worth it.

But that's the.

(Josh Tyrangiel) Okay.

But what about the competitive set though.

(Reid Hoffman) Oh, and then pharma.

Yes.

So, look, look, all companies learn how to be very professional at the thing they're doing.

But tend to be bad at new technological trends.

And sometimes those companies themselves, you know, embrace and extend the new technological trend.

So for example, Microsoft's doing very well at AI, but by the way, didn't do so well at mobile.

Right?

So it's like even, the, the most elite, most competent tech companies, you don't always hi that, that curve the right way.

Sometimes you do there.

Everyone else, the new technology is usually like they just don't know how to build and embrace.

Pharma is the same way.

So in terms of their biological science, in terms of their understanding of the the kind of, the regulator process, the therapeutic process of, you know, world class AI, don't understand it at all.

(Josh Tyrangiel) So it's sort of hammers and nails.

(Reid Hoffman) Yeah.

(Josh Tyrangiel) And, and is your suspicion, I mean, obviously you're making a bet.

Is your suspicio that AI is transformative enough that approaching it from the other side will yield gains faster, mor furious than, than they can do?

That is 1,000% the bet.

And I'm certain the bet is true.

The question is, can we do it?

(Josh Tyrangiel) Got it.

So, you work very closely with Microsoft.

You're on the board.

And they are giving you some tools as well?

(Reid Hoffman) Oh, yeah.

So I mean, looking.

(Josh Tyrangiel) Yeah.

Describe some of those tools and let us know a little bit what may not be on the market now that you might get to use in this fight.

(Reid Hoffman) Well, so one of the things that, Microsoft has is, is kin of Microsoft research, which is, a whole bunch of extremely smart people who basically are paid to kind of just do research and they research on a number of things, including quantum.

So there's a quantum program that's come out of out of Microsoft.

And, and some of that is, by the way because there's a natural thing between software and now AI and biology, like, it's like some of the world's best computational chemists.

And like, I wasn't even aware that there was a discipline of computational chemistry until we started going down this path.

And then since, like, well, you know, there's this guy who works at MSR, I'm like, well, we can go meet with MSR.

That's very easy.

Right?

And so and so the, there' a number of different of the MSR folks have, have constructed very impressive computational chemistry tools.

And it's again, part of the thing that we're doing at Manas, is that we're not trying to to only say we're going to build all of it ourselves.

If someone else has a piece of technology that's really helpful, we're just going to use it.

Right?

(Josh Tyrangiel) You touched on this a little bit, but I want to, project forward.

You know, it's ten years out.

Someone's diagnosed with stage two lung cancer, which means that that, tumo of a four centimeters or less.

We know what would happen today.

Why don't you just take a step by step?

What would happen ten year from now if you're successful?

And how that's different.

(Reid Hoffman) Well, my, hope would be, especially for something major like lung cancer, that we would have discovered a molecule that, for that particular cance would literally, all you would essentially need to d is get it into your bloodstream and would bind with the righ cells, kill those cells, right.

And nothing else.

And, you know, obviously the, the, the gold standard is could it be something you a pill you swallow.

If it's an injection, fine.

Right.

And if it's an injection that has to be injected in a specific place, you have to go to a clinic.

Fine.

These are a million times better than chemo.

Right?

So, you know, and by the way, of course, if it's still super small, then you might still do a surgery, right?

It depends a little bit on what the efficacy is are.

But like it's it's kind of the question of where you, where you, where you fit between, you know, on that spectrum.

But once you get out of the, the surgery thing, the idea would be a molecule that essentially solves your cancer.

(Josh Tyrnagiel) And you think that's feasible in ten years?

Less?

(Reid Hoffman) It's certainly feasible in ten years.

I mean, again, can we do it?

But but it's.

(Josh Tyrangiel) But someone will do it.

(Reid Hoffman) Yes.

(Josh Tyrangiel) Okay.

So I mentioned you live this kind of portfolio life, which is great.

One of the joys of which is you have really spent a lot of time with AI been an enthusiast, what's change the most in the last six months?

Because you've had a lot of pretty strongly held feelings, but this thing is a roller coaster.

So where have you shifted in the last six months?

(Reid Hoffman) Well, I continue to be super positive, as you know, because the super agency and all the rest and we've talked about that before.

I'd say that, probably let's see, the key things to tracking updates are, like this year a bunch of the Chinese efforts have demonstrate that they are actually, in fact, very much in the game, that there is a set of, of different things that they have now.

There were a bunch of spurious claims like, oh, we could train a frontier mode for much cheaper than you are.

And that's just all fiction.

But the but, but they're trainin good models.

And there's at least four Chinese models that I'm trackin right now as, as, as interest, having interesting capabilities and things you do.

That's one.

Two is, it was started earlier than this, but it's one of the things that most people in general audience here tracking is, how AI's reasoning capabilities are improving and if, for example, speaking to everyone here and online or whatever else, if you haven't tried deep research on ChatGPT or Gemini or Claude or, you know, pick your favorite frontier model, do.

And what specifically is to do like, is you want to it's, it's the kind of instruction you would give, kind of like a research assistant.

Like he'd say, okay, I'd like you to, you know, analysis, I'd like to ge a detailed analysis of the last 30 years of what advancements have been made in cancer medicine.

And, and I would like to include these particular things.

And, you know, and then it will go and compute for, call it ten minutes, and then come back wit something that's pretty amazing.

Doesn't mean it's always, it doesn't mean that aren't sometimes hallucinations and other things, but it's, it's like stunning how good it is.

And it's getting better each time.

(Josh Tyrangiel) Yeah.

And I would add, just for people who haven't used it, as a journalist, it is very easy to put it in, include citations and then fact check those citations.

So the the hallucination problem, which is still real, is, much less problemati than it was even six months ago.

(Reid Hoffman) Yeah.

But, but by the way, so the chain of thought thing, the reasoning, the reason I mentioned it is, is that reasoning capability is another vector that's that's increasing improvement every quarter.

And so part of the thing that, you know, the journalist cycle is, you know, likes to go, oh my God, it's going to be everything.

Oh my God it's oversold.

Oh my God.

It's going to be everything.

Oh my God it's oversold.

It's not actually oversold, right?

Because like there's vectors for improvement on it.

And then the very last one is a way to think about kind of what the futur you're going to be living in is, is we are all going to, anytime we're solving any serious problem, deploy one or more agents.

And what's more, in the creation of software, all of us will be in some number of years be creating custom software for the things we want.

(Josh Tyrangiel) Does your optimism extend to the transition in the labor markets?

(Reif Hoffman) Long term, yes.

Short medium term, it's transition is massiv difficulties with transitions.

Like I anticipate massive difficulties.

The reason I say, as you know, cognitive industrial revolution is both the outcom being really, really important.

But the transition the, the transitions that happen being also quite difficult.

(Josh Tyrangiel) So one of the challenges is actually just measuring where we are in that transition and what's happening.

If you were Secretary of Labor, what would you be looking at to get early signals?

And if there weren't the right signals, what would you fund or create as a metric that would help us know what's going on?

(Reid Hoffman) Well, it's a good question.

You know, the, tempted to say something snarky about the current administration, but I will pass on that.

(Josh Tyrangiel) Oh, no.

Yeah, because I have other opportunities.

Let's just.

(Reid Hoffman) So, but the I guess the thin that I would be looking for is.

Okay, so you you want to take a look at, the key place where you will see a lot of different transitions are two forms.

One is, where will do the jobs of replacement, be customer service, etc., will be part of that.

And then where will the jobs of application be?

And then what happens i the patterns of amplification.

Now, in the first patterns of amplification, that will actually also look like job reduction.

Because as the transition goes and you want peopl who are doing the amplification, but actually, in fact, I'm thinking a lot of those cases that will be a temporary effect versus a enduring effect, because say, for example, you know, you and I are compan one and company two, and we are, competing with each other.

And marketing i one of the things we compete on.

Sure.

Maybe, like my current our current marketing department is at 20 people could now do that same way.

Could work with four people, but we're competing with each other.

And so part of ho we got the 20 people right now is because we're still competing with each other and doing it.

(Josh Tyrangiel) Right.

(Reid Hoffman) So we'll probably get back to a different set of configuration of 20 people and competing.

But that's the, that's the kind of the amplification story.

And there's a lot of places where it actually has that initial kind of dip, and then figuring ou how you can figure whether it's product development, because actually I think there's infinite demand for creating, you know, new software and product development.

So the whole like, oh, software engineers are going away, I think is is literally just like silly, right.

But customer service, I don't think that the same number of customer service jobs will exist on the other side.

And what's more like one of the things that I find is kind of entertaining.

And I will get back t your Secretary of Labor question is kind of entertaining is one of the things we're seeing is people are calling a human, has a human on their side an say, you know, put the human on.

Right, because they're so frustrated with the huma who's trying to follow a script that they think it's a robot.

(Josh Tyrangiel) Yes.

(Reid Hoffman) Right.

So, so it's like, okay, that's part of the reason why I think, like the very first call, but that will transform what is customer service experience, because right now it's how do we hire someone for at least possible wage anywhere in the world, have them follow a script and get you off the phone is essentially the customer service thing.

Now, if you're running AI, it's how do I help you?

Hey, are you having a good day?

Is there anything else you to like?

Its fine, right, to do all that stuff.

So it'll be transformed experience, but there will be jobs.

So now on the Secretar of Labor side, it would be okay.

Take a look at thes two kind of categories of jobs and then be tracking wha current volume both entry level, Stanford did some really good work here.

You know energy coming in and current volume is and then if there are, repeated changes going on, think about like what kinds of things you want to do to try to help with job transitions.

And actually getting a set of, you know, kind of good thinking and ideas because one of the things we know is importan for the functioning of society.

It's okay if jobs changes.

Okay if you like the, you know, I know there's still horse and buggy drivers here in New York.

(Josh Tyrangiel) Yes.

And I still have a I still have my footman and he's great.

(Reid Hoffman) Yes.

(Josh Tyraniel) But yeah.

(Reid Hoffman) But you know, like, not so much anymore.

(Josh Tyrangiel) Yeah.

So that that really leads into the next question, which is that, for better or worse, we currently live in a civilization.

And that civilization has different sectors that have to collaborate i order to manage these changes.

(Reid Hoffman) Well, you're reminding me of Mahatma Gandhi' quote on Western civilization.

It would be a good idea.

(Josh Tyrangiel) It's true.

So let's say one day we get one.

Where are you on the role of politic in helping with this transition?

Do you have any faith in th current political environment?

Can we accomplish something as large as transitioning a significant sector of the workforce through this uncertainty?

(Reid Hoffman) Well, I think obviously, you know, you know, I worked pretty hard to try to not have the current administration, and, you know I have a whole list of things, whether it's from, you know, insanity around vaccines to, you know, kind of challenges around, you know, like tariff and volatility and what it does for business markets and prices and everyday living and jobs.

And also it's like we could spend three hours just listing all the stuff so that intrinsically makes you less optimistic.

Now that being said, I think that the, the responsibility for all citizens is to try to make it work and so, you know, that's part of the reason why I went, okay, you know, this, this administration doesn't want to hear advice from me.

That's fine.

I will go build stuff and make it happen.

But I think that, that, you know, we are not, look, if we're having a discussion about whether or not vaccines make sense, right?

Or, you know, other thing where there is extremely clear scientific evidence on these things, it makes it much harder to have fait that other kinds of good, call i intelligent governance, play out well.

(Josh Tyrangiel) So, you have.

(Reid Hoffman) I omitted all the expletives.

(Josh Tyrangiel) Yes.

I thought, I thought you did a great job there.

The conversation aroun AI and safety has shifted fairly dramatically and in large part, you think I'm pivoting off this subject, I'm not.

In part because we now understand the power of authoritarian regimes in AI.

So let me ask you a bit of a hypothetical that I hope we don't, have as a practical, in what ways can an authoritarian government use AI currently, as currently exists in the marketplace, to increase its control over society?

And how would we know that they're using it?

(Reid Hoffman Oh, the second part is a little bit more challenging.

Although part of the question okay.

So the first one is it's already happening.

It's happening in China.

There's, there's, there is like, you know, a bunch of things in terms of Uighurs and other things that it's already happenin that I've, years back, not many, not within the ten year window, but a, an investor a friend of mine went to a company in, in Beijing and for an investor pitch, and they showed him here's where everywhere you've been in Beijing because they had all the the footage and he's like okay, thanks, I'm leaving now.

And so, so that's like there's, there's a number of different things that could easily be done now and are being done in, in some places now tracking that it is happening, I think is probably mostly a function of, you know, well-ordered societies, which we're having some trouble with right now.

The, there's, the most cutting edge you have to be working with the companies and one of the things that the companies are good is they have alternative, like they have a board of directors, and their shareholders and a bunch of other things.

And most of those companies, you know, are pretty explicit about, we refuse to have our technology used for mass surveillance of citizens.

Right.

And (Josh Tyrangiel) So far, yes.

(Reid Hoffman) Yes.

(Josh Tyrangiel) Okay.

We're going to end it o a slightly more optimistic note.

You are famously networked.

This is a networking event of some kind.

I'm curious, what conversation have you had over the last year or so that's just completely blown your mind?

Change the way you think about the world.

(Reid Hoffman Well, probably the good news is I probably encounter those reasonably often.

That's, that, that's the the one of the things I love about, you know, the kind of place where I end up.

But maybe this would be the kind of, because it's something like I know really well, but I still learn something really amazing.

So, I was sitting with Ethan Mollick at Wharton, who is like one of the great, like, really, kind of leading thinkers and, like, follow his dreams and so forth.

And he was like, oh, people are undercutting.

They don't understand all the stuff that you can already do with multimodal AI.

And I was like, sure, I think I know everything.

Tell me what you're what you're what you're thinking.

And he said, well, one of the the big problems in construction projects is it's very hard to know if you're on track if there's a long pole, tent if if, if, if there's something can do to try to accelerate it.

All the rest.

And so you have inspectors go by, but the inspectors are like one time slice and very expensive do.

And, and, you know, kind of manage all the rest.

So what he did is on a construction project, he set up 24 cameras.

He loaded into the AI the construction plan, and they asked the AI to do a report at each day where the construction was, because it was reading in multimodal for the cameras and got a detailed like, and this is like a guy who's not even really a coder got a detailed like, here's the things that are working well, here's the things that are at risk.

Here's the things that are falling behind.

Here's the things you would need to do in order to try to stay on schedule, etc.... (Josh Tyrangiel) Wow.

(Reid Hoffman) And at the end, th his contractor was still like, you can have it fast, you can have it cheap.

Yes, exactly.

(Josh Tyrangiel) Goal, AI.

(Reid Hoffman) Yes, yes.

(Josh Tyrangiel) All right.

Reid, always a pleasure.

Thank you so much for being here.

[applause] [instrumental music]