NPR's guide maps six factions inside the AI safety debate

The debate over AI safety is not the two-sided fight it looks like from outside, NPR reports. Katie McQue's guide sorts it into six factions with overlapping and sometimes contradictory goals.

The hook is a September 8 resignation. Jacob Coxon, a researcher who had worked at OpenAI and then Anthropic, walked out of Anthropic and publicly accused the companies of "racing straight to self-improving superintelligence and gambling with our lives."

The six camps McQue lays out: the effective accelerationists and tech right, whose named figures include White House AI and crypto czar David Sacks, venture capitalists Marc Andreessen and Ben Horowitz, OpenAI president Greg Brockman and Nvidia CEO Jensen Huang; a populist right skeptical of Big Tech that pulls in former Trump strategist Steve Bannon and former Fox News host Tucker Carlson, joined from the left by independent Vermont Senator Bernie Sanders; the "safetyists" worried about superintelligence, associated with Geoffrey Hinton and former OpenAI researcher Daniel Kokotajlo; the effective altruists, whose figures include MIRI co-founder Eliezer Yudkowsky, Anthropic co-founder and CEO Dario Amodei, and FTX founder Sam Bankman-Fried, convicted in 2023; the AI ethics scholars, associated with former Google AI research scientist Timnit Gebru; and pragmatists who treat AI as normal technology, associated with Stanford Human-Centered AI Institute co-director Fei-Fei Li.

The factions cut across the usual political axes. At a Pro-Human Assembly conference, Bannon told the room: "We should quarantine now every aspect of the ecosystem of artificial intelligence away from the Chinese Communist Party today." The Defense Department, meanwhile, posted the line "Americanism, not effective altruism. The United States will continue to be AI DOMINANT!", a swipe naming one of McQue's six factions directly.

By the time we picked it up, two researchers we track had already shared it.