Reid Hoffman, the co-founder of LinkedIn, is in a difficult position — on the outs with President Trump, a new administration and his former friend Elon Musk.

Hoffman supported Kamala Harris for president in 2024, and had previously bankrolled a lawsuit against Trump by E. Jean Carroll, who accused Trump of sexual assault. And on January 20, when several tech billionaires, including his friend Mark Zuckerberg, stood on stage for Trump’s inauguration, Hoffman wasn’t invited. He told us that he assumed it’s partly that he isn’t a current CEO of a giant company, but the political difference with his fellow tech founders is hard to overlook.

We had a conversation with Hoffman about this awkward moment — and also his effort to look far beyond this moment, into the future of artificial intelligence. Hoffman, a longtime investor in AI, has co-authored a book called “Superagency” that takes an optimistic view of AI. At a moment when many Americans worry about what can go wrong with AI, Hoffman argues for what can go right.

Here are four highlights from our talk, which you can listen to at the link above.

He used AI as a sounding board and fact-checker.

Hoffman says he and his co-author Greg Beato did not ask a chatbot to write the book, but used a chatbot to vet the book — checking facts, or asking for additional research, readings and perspectives on artificial intelligence. (He said he also asked for help finding good books on AI, though he contends the technology is developing so quickly that “there aren’t any good books.”)

This illustrates Hoffman’s case for how AI can be helpful — as a superpowered information sifter, which can help human beings track, locate, boil down and analyze the infinite amount of information available on the internet.

He has an explanation for his falling out with Elon Musk.

Hoffman insists he lost a friend over Musk’s political pronouncements on X, his social media platform. “While I really highly regard Elon’s amazing accomplishments in cars and rockets and all the rest,” he has been “using Twitter” to spread “lies and slanders.” Hoffman also alluded to their differences of opinion over OpenAI, a company in which both invested.

He is trying to keep his friendship with Mark Zuckerberg.

Zuckerberg critics accused him of bowing before Trump, or blowing with the wind, by abandoning professional fact-checking on Facebook in the U.S. and attending the inauguration. Hoffman defends his friend—as a man who is “trying to build a social infrastructure for the world,” and “believes very deeply in the freedom of speech,” but who should be “responsive” to governments as a responsible CEO.

Hoffman still said he believes in fact-checking, however. “Social networks should, in fact, you know, kind of take a certain social responsibility.”

Hoffman tries to make the best of his fellow CEO’s on stage with Trump.

“If I were to pick one industry that I would most want to be in dialogue with the president’s office and with the American government, it’s the tech industry.”

Does he fear the new administration will seek retribution against him? “My hope is the focus of the administration will be on: How do we build as much strength into American society, American industry as we can [as opposed to] claims of retribution. I don’t think it’s that useful to speculate right now. We’re in the administration. We’ll, we will see what happens.”

Transcript:

STEVE INSKEEP, BYLINE: Reid Hoffman has been thinking about how artificial intelligence can help him think. The cofounder of LinkedIn, the social network for business, has invested in AI for years. I asked him if investors have created an AI bubble, overinvestment, and he said no.

REID HOFFMAN: If anything, we’re as yet still underinvested and under-rotated towards this kind of new artificial intelligence transformation.

INSKEEP: Underinvested? You mean we need a few more trillion dollars thrown in this direction?

HOFFMAN: (Laughter) We aren’t at trillions yet. We are getting there. I think we’re just at the very beginning of this really incredible – what I call the cognitive industrial revolution.

INSKEEP: Hoffman recorded this last week, before the announcement of a cheap Chinese chatbot sent some U.S. tech stocks sliding. In a new book called “Superagency,” Hoffman asserts that AI will empower people. He says it will help – in fact, already helps – to sort the infinite amount of information on the internet.

HOFFMAN: And obviously, some of it’s very good, and some of it’s very bad. And so how do we navigate that? How do we figure out which things are accurate? And sometimes it’s really important medical information. Sometimes it’s important because it’s information about what’s going on in the world. And sometimes it’s, you know, is this going to be a good recipe or a bad recipe for dinner?

INSKEEP: The idea is that AI helps me process infinite amounts of information. But it seems to me likely also that AI is going to just keep producing more and more information. The science fiction writer Ted Chiang wrote an interesting article in which he pointed out how likely it is – and I bet it already happens – that some executive is supposed to write a 50-page report, so he gives a few prompts to a large language model, which generates a 50-page report, which gets sent to its recipient, who never reads it. He just has his chatbot boil it back down into bullet points. Like, what is the point of that even?

HOFFMAN: Well, it’s definitely probably happened already and going to happen more. And obviously, if I’m responsible when I’m producing information, I might use an AI system to help produce it versus type it out all myself, but I’ll read it, refine it, make sure it’s very accurate. And then you have both the long form, which is what I’m sending. And then the short form, which I might say, hey, I don’t need to look at the long form. All I need to do is see the short form. Hey, I don’t have time to read the book “Superagency.” I’d really just like to have a three-page precis.

INSKEEP: (Laughter).

HOFFMAN: Could you please give me the three-page precis?

INSKEEP: Or a short interview on NPR, just to give another hypothetical situation. But go on, please.

HOFFMAN: (Laughter) Yes, exactly. But it’s a feature, not a bug. What’s relevant is, can we find the stuff and engage with the stuff that really matters to our lives and to our work?

INSKEEP: Did AI play any role in the writing of this book?

HOFFMAN: It did. We didn’t have AI write the book. We double-checked things. So we’d say, hey, we’re writing about what the role of the printing press was or what was happening with cars or mainframes. And then we would go to AI and say, OK, how would a historian of technology think about this? What would their critical analysis of it be? Is there anything we missed?

INSKEEP: So you hired a fact-checker, and the fact-checker was a large language model.

HOFFMAN: Exactly.

INSKEEP: Reid Hoffman is like a lot of tech leaders in his enthusiasm for AI. In other ways, he’s walked a different path. In 2024, as Elon Musk and others supported Donald Trump, Hoffman backed Kamala Harris. He’d even bankrolled a lawsuit against Trump. And then Trump won, and Hoffman saw many of his tech peers attend the inauguration and sit onstage with the president he opposed.

HOFFMAN: If I were to pick one industry that I would most want to be in dialogue with the president’s office and with the American government, it’s the tech industry.

INSKEEP: It sounds like you approve of these tech leaders showing up on the inaugural stage, even though you were so politically opposed to this particular president.

HOFFMAN: Well, you know, in ’24, I obviously made a great effort to, you know, what I think, you know, would’ve been an alternative choice. But part of what I think the responsibility of all citizens, of which I obviously count myself as one, is how do we build the best possible future for Americans? There’s obviously going to be things that I have deep questions about, you know, the January 6 pardons, other kinds of things. But building the technological future is, I think, really important, and so you should call out the things that I think are positive.

INSKEEP: Critics of Mark Zuckerberg, who I believe is a friend of yours, said a variety of things about Zuckerberg’s about-face on policy, as well as showing up at the inauguration, that he flipped politically, that he bent the knee, that he’s blowing with the wind, that he’s going for the money. What do you make of those critiques?

HOFFMAN: Well, you know, I myself have been a voice that I think social networks should, in fact, take social responsibility. I think fact-checking is important.

INSKEEP: Which Facebook abandoned. Right, go on.

HOFFMAN: Yeah. And so, you know, I have a different point of view. A lot of that’s reflected on LinkedIn, you know, the social network that I cofounded. The key kind of, like, meta, you know, question – you know, maybe pun intended here…

INSKEEP: (Laughter) Well done. Please, proceed.

HOFFMAN: Exactly. What we need to get to as a society is, you know, how do we get good markers for discovering truth?

INSKEEP: Do you think Zuckerberg is blowing with the wind?

HOFFMAN: I think – let’s see. I guess what I would say is he’s very much trying to build a social infrastructure for the world. He believes very deeply in the freedom of speech, so I don’t think he’s blowing in the wind in any of those contexts, you know, because I do think that technology companies should be responsive to their governments. Is he being responsive to that? Yes. But, you know, I think that’s part of what, as a business leader in any business, you should be doing.

INSKEEP: Why did you fall out with Elon Musk?

HOFFMAN: Well, I think, unfortunately, while I really highly regard Elon’s amazing accomplishments in cars and rockets and all the rest, you know, he’s been basically using Twitter to kind of say a bunch of things that are kind of lies and slanders. There’s no basis other than maybe competition with AI, maybe, you know, unhappiness about the fact that he made some of the wrong choices around his own involvement in OpenAI. But that’s all speculation. All I see is the tweets that have, you know, kind of slanders and lies and no information in them.

INSKEEP: There’s no personal backstory?

HOFFMAN: No.

INSKEEP: It’s just his public behavior has turned you off?

HOFFMAN: Yeah. Yeah.

INSKEEP: Having contributed to Kamala Harris’ campaign, having been so outspoken, having bankrolled a lawsuit against President Trump, what are your concerns, if any, about retribution?

HOFFMAN: Well, my hope is the focus of the administration will be on how do we build as much strength into American society, American industry as we can, versus the various, you know, claims of retribution. I don’t think it’s that useful to speculate right now. We’re in the administration. We will see what happens.

INSKEEP: Reid Hoffman, it’s a pleasure talking with you. Thank you so much.

HOFFMAN: Likewise. Have a great day.

INSKEEP: His book is “Superagency.”

(SOUNDBITE OF MASAYOSHI FUJITA’S “MOROCCO”)