Confidence Men

I've written already about some of the things I enjoyed most about June's book of the month, Empire of AI, and hopefully those of you in the club have your copies and have dug in by now. You will see that it's meticulous, opinionated and wide-ranging—three great qualities.

You might want to know more about how the book was built, or what the author thinks about the latest developments in AI—and so I want to remind you that this Thursday we have a live interview and Q&A with author Karen Hao, where we'll be discussing all this and more.

Everyone on our mailing list has an invitation to join the Zoom at 8pm Eastern, 5pm Pacific on Thursday.

If you have questions to ask Karen, you’ll be able to ask direct in the room or you can email me in advance via [email protected].

⌘

However, Empire of AI is far from the only book about artificial intelligence and the people behind it right now. It's not even the only book about OpenAI released in the last month. So I thought I'd tell you about two other relevant books I've read recently so you can choose whether you want to pick them up or not.

The obvious place to start is Keach Hagey's The Optimist: Sam Altman, OpenAI, and the Race To Invent The Future. It's the most direct parallel to Hao's book, because it revolves around the same central people. But they're very different pieces of work. Like Hao's book, The Optimist tries to tell the story of what's going on behind the scenes of this seemingly unstoppable technological advance. Unlike Hao, though, Hagey relies almost entirely on Altman's origin story as its lens, which leaves far less time and space to look at the consequences of what he’s doing.

Part of this may be because it's chronological, describing the steps in Altman's rise in painstaking detail. It's almost 200 pages in before OpenAI actually starts to elbow its way into the narrative. But it's also because Hagey tries to buttress this detailed recounting by mining every incident for meaning. Small moments are constantly fed to us as if they are meant to foreshadow greatness—looking in Altman's history for clues to his dealmaking prowess, say, or searching for proof of his exceptionalism everywhere (the fact he could play a videotape in the VCR at two is written portentously.)

She also, somewhat oddly, dwells on physical appearance, often describing people in ways that just don't jibe with my own knowledge. For example, startup investor Paul Graham "doesn't look like a nerd"—when in fact, he looks like an IBM middle manager on the weekend. Others get similarly odd treatment. And Altman himself is constantly described as a kind of preternaturally youthful imp—a small "elfin", "dimpled", "rail-thin" man—when in fact he is of average height, a little bobble-headed, with watery eyes. I wonder why physical characteristics are given almost moral weight.

The Optimist takes great pains to stress that Altman "did not want this book to be written"… until the moment that he did. At that point, Hagey notes that Altman cooperated with hours of interviews. His cooperation does seem to have soft-soaped many of the strongest criticisms—for example, the self-enriching double-dealing outlined in Hao's book is mentioned in passing; the externalities of training huge LLM models barely get a mention.

Both books do paint a similar picture of Altman as a person: one who projects wild confidence and will do almost anything it takes to make a deal, including overpromise; one who is intensely driven and fiercely competitive but averse to conflict; one whose positive perspective of technological progress has become quasi-religious over time. Above all they portray a person who is drawn towards power; first being near it, and then having it.

The Optimist is a well-reported, newsy biography, and Hagey appears to have done a prodigious number of interviews. But ultimately it ends up feeling thin, a solid tick-tock of how we got to now that's far more interested in how Altman got to where he is today than with the real-world implications of what he, and OpenAI, is doing to our future Hagey is much more interested in mythmaking than in examining what OpenAI is doing… or what its principals actually believe.

⌘

If you want to know more about those beliefs, then I would highly recommend More, Everything, Forever, Adam Becker's seething, scabrous examination of the mental frameworks that Silicon Valley's elite hold dear. (Full title: More, Everything, Forever: AI Overlords, Space Empire and Silicon Valley’s Crusade To Control The Fate of Humanity.)

Altman is in there, along with various AI thinkers, doomers, boomers, and technological ultracapitalists: Elon Musk, Jeff Bezos, Ray Kurzweil, Nick Bostrom and others all feature. Each of these individuals holds a set of intertwining, overlapping beliefs that bounce off and inform each other; viewpoints that twist and turn and nest and lean on one another. Between them they create an almost-religious framework that must seem logical from the inside but looks fragile and self-aggrandizing from an outside perspective.

These beliefs include rationalism, transhumanism, space colonization, the singularity, longevity and effective altruism. Each of these gets a solid chapter, and while not every person holds every one of these beliefs, Becker shows how their entangled mesh is built brick by brick… and promptly demolishes these walls of ideas.

An astrophysicist by training, Becker brings scientific clarity and learned precision to his arguments, while never straying into unreadable academic prose. What do these people believe and why? He lays it bare. There's a lot of magical mathematics and fallacious leaps of logic, a kind of underpants gnome syndrome that is often in total contradiction to physics. He shows that when scale is your god, and "we'll be smart enough to solve this in the future" is your hedge against today's externalities, that an apocalypse caused by a superintelligent computer and an apocalypse caused by you only seem marginally different.

⌘

Perhaps these three books are best read as a triptych. After all, many of the characters and moments in More, Everything Forever are shared with The Optimist and Empire of AI we've mentioned: Hagey and Becker both feature AI doomer Elizier Yudkowsky prominently in their opening chapters; folks in Hao's book pop up with regularity in the others. And they all chase down many of the same ideas from different directions, wrestling with extropians, colonization, seasteading and more all pop up over and over.

Hagey's focus on Altman gets you furthest inside his particular personal story, if that’s what you’re interested in; Becker does the most to interrogate what Altman and his cohort actually think; Hao does the best at tying all these together and showing the bigger picture of how the world changes and what could be different.

But read all three and you'll see not just how deeply embedded this set of ideas is in Silicon Valley's elite, but how deeply immoral some of it is too.

Onwards

Bobbie