Hi everyone — well, this is embarrassing. 💀 I was using AI to research and test an SEO page, and somehow managed to send it to my entire email list. THIS is the article I actually meant to send. It’s part of my new series exploring the people behind the headlines. (And for the record, no,, I do not use AI to write my articles, but I do reserve it to help me make the most egregious mistakes hah) I hope you enjoy!
—
Anthropic has officially filed its S-1 with the SEC for an initial public offering, which means it’s poised to become the first frontier AI lab to go public.
Anthropic has raised approximately $141.1 billion in capital since its founding, according to its S-1. Its largest outside shareholders are Amazon and Google, which will each own more than 5% of the company after the IPO, all of it in non-voting shares.
One of the risk factors Anthropic lists in its S-1: “Navigate an evolving, uncertain, and at times divergent global regulatory landscape and engage with legislators, regulators, and other stakeholders.” (Allow me to translate: “We’re building technology faster than governments can figure out what the hell it is, and now we have to explain it to a bunch of politicians who still struggle to open a PDF.”)
Now, to the part I found most interesting. Buried on page 145 is the name of a man I’d never heard of who wields an unusual amount of power: Dr. Neil “Buddy” Shah, chair of the Anthropic Long-Term Benefit Trust.
Shah’s job is essentially to help protect Anthropic from its own worst instincts. He doesn’t work at the company. He owns no stock in it, and the rules of his position prohibit him from buying any. He’s compensated in cash, with no financial upside tied to Anthropic’s success.
Since June 2022, he’s led the Clinton Health Access Initiative, which helped negotiate access to generic versions of the HIV-prevention drug lenacapavir for roughly $40 per patient per year, compared with a U.S. price of about $28,000 annually, according to Shah.
In Anthropic’s IPO prospectus, he’s listed alongside co-founders Dario Amodei, Daniela Amodei, and Jack Clark as one of four initial managers of a new Public Benefit Fund, which is slated to hold 5% of the company.
As chair of Anthropic’s Long-Term Benefit Trust, he holds a special class of stock called Class T. Class T shares carry no economic rights. Instead, they give the trust the power to elect and remove a majority of Anthropic’s board of directors.
Basically, Shah helps oversee a trust that can decide who sits on the board of one of the world’s most powerful AI companies. Shah has chaired the trust since June 2023.
He previously explained on a podcast that because AI could eventually end up “within the hands of a few people,” Anthropic’s founders wanted to “tie our hands to the mast” and give “meaningful corporate governance control to a set of non-financially interested trustees.”
We gotta talk about how unusual of a setup this is. Most founders spend their careers fighting to retain control of their companies by creating dual-class shares, negotiating board seats, and structuring voting rights in a way that make sure no one can tell them what to do.
Anthropic’s founders, on the other hand, have given a group of outsiders the power to help decide who gets to oversee them. (To be clear, the founders still retain 50.1% of the voting power on general shareholder matters through a special class of stock. But the trust gets to pick a majority of the board.)
I find this reassuring and also sliiiiightly terrifying. It’s reassuring because at least someone at Anthropic seems to recognize that what’s best for shareholders isn’t necessarily what’s best for the rest of us. And terrifying because what exactly do the founders know about the tech they’re building that makes them believe they need this sort of protection from themselves?
So the question is: Who is the person getting to decide what’s in humanity’s best interest? Well, Anthropic has entrusted some of that power to a handful of people most of us have never heard of. Buddy Shah is one of them.
But having no financial stake in a company doesn't automatically mean you have better judgment. You can own no stock & have no financial conflicts and still make the wrong call.
No disrespect to Buddy, but the only thing more dangerous than an AI that thinks it knows what’s best for humanity may be a human who thinks he does.
A “Shah” in a position of trust. Look for the scam.