风险大佬论OpenAI人事地震

In the land of the acceleratonists and the doomers, the techno-optimist stands tall. As Silicon Valley was roiled the weekend before Thanksgiving over the sudden firing of the face of AI, Sam Altman, the billionaire venture capitalist Marc Andreessen was typically vocal on Twitter, saying the equivalent of “I told you so.” It just looked a lot more like “e/acc,” instead.

The founder of Andreessen Horowitz, who has staked his claim firmly in the “accelerationist” camp for years before AI dominated the business climate of 2023, made his stance ultra-clear in October. His 5,200-word “techno-optimist manifesto,” posted on both his own substack and Andreessen’s blog page, ripped into skeptics of big tech in 15 sections, often lyrical and poem-like. The conclusion is there right at the start: “We are being lied to,” he writes, excoriating those who claim technology “takes our jobs, reduces our wages, increases inequality … [and] degrades our society.” (Andreessen has been writing tech-boosting manifestos for years before AI exploded on the scene, such as 2011’s “Why software is eating the world” and 2020’s “It’s time to build.”)

It’s still not publicly known exactly why OpenAI, the most pivotal firm in the development of generative artificial intelligence, ousted its cofounder and leader Sam Altman, but multiple reports point to exactly these kind of worries as the reason. The “accelerationist” camp that welcomes AI for being, as Altman said onstage in San Francisco just last week, like a Star Trek computer, is opposed by the “doomers,” who worry about it posing an existential risk to humanity. Elon Musk surprisingly leads their ranks, and he’s criticized OpenAI from straying too far from its nonprofit mission to benefit humanity since he left in 2018.

Understanding the nature of OpenAI’s founding is pivotal to understanding the developing story of Altman and his board. It was co-founded in 2015 by Altman, Musk, Ilya Sutskever (who remains at OpenAI and was one of the board members who ousted Altman) and Greg Brockman (the former board chairman who was stripped of his status, before resigning in solidarity with Altman). They explicitly chose to establish OpenAI as a non-profit entity in response to Google’s acquisition the year before of DeepMind for $600 million, and their fears that Google’s lead on AI would grow into a potential monopoly. Developing AI for the benefit of the humanity was always crucial to OpenAI’s mission and the non-profit board and structure sits atop an unusual capped-profit subsidiary. The “accelerationist vs. doomer” tension was therefore embedded into OpenAI’s structure before exploding this past week.

e/acc!

Andreessen’s activity over the weekend confirms that he sees the issue in terms of accelerationists versus doomers. “E/acc!” he posted on November 18 afternoon, a shorthand for “effective accelerationism.” It’s a play on the theory of “effective altruism,” made infamous by disgraced crypto fraudster Sam Bankman-Fried. As explained by Business Insider, the ideas of effective accelerationism likely originate from Nick Land, a British philosopher credited as the father of the broad accelerationism movement, but have since grown into a belief that innovation and capitalism should be maximally exploited to achieve radical social change, even at the expense of social stability in the present.

What Andreessen chose to retweet spoke even more explicitly in this direction. At one point, he reposted a thread tracing back to Matt Parlmer, a founder and CEO of a Michigan-based company called General Fabrication. “Doomer people,” Parlmer wrote, obliquely referring to Altman’s firing, “I hope you all recognize that you shot your shot with this ridiculous stunt and now everybody is ready to bounce you out of polite sane person company for quite some time.”

Parlmer’s post prompted another reply from an account called “Beff Jezos — e/acc,” who has a substack dedicated to effective accelerationism. “This was always their plan,” wrote their reply. “Infiltrate. Subvert. Co-opt. They spent their one shot. Now no one will trust EAs [effective altruists] and Doomers ever again. Make sure your board is e/acc-friendly rather than EA-inclined, anons.”

This sentiment was widespread in Silicon Valley posting over the weekend, recalling the febrile climate earlier in the year when Silicon Valley Bank suddenly failed. David Sacks, the prominent venture capitalist and Elon Musk ally, who was likewise vocal during the SVB collapse, seemed to speak for the Valley getting tired of Doomers in general and nonprofits in particular. “Sam should get his job back, the board should be replaced by founders and investors who have skin in the game, the nonprofit should be converted to a C-corp, and Elon should get shares for putting in the first $40M+. In other words, undo all the shenanigans.”