From Mother Jones · Satchel Walton · · 1m 10s
AI Safety Community Fears 2027 Intelligence Explosion
AI safety experts warn of imminent existential risks from AI as development accelerates.
In brief
AI safety experts warn of imminent existential risks from AI as development accelerates. AI safety researchers fear a 2027 intelligence explosion could cause mass casualties or government overthrow within months. Originally reported by Mother Jones.
AI Safety Community Fears 2027 Explosion

AI safety experts warn of imminent existential risks from AI as development accelerates.
“Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all...i am scared, sad, and angry”
2027 Intelligence Explosion Timeline
Daniel Kokotajlo predicts 2027 intelligence explosion with mass casualties or government overthrow within months.
“i am scared, sad, and angry”
OpenAI and Anthropic Speedup Reports
METR reports Anthropic's ai research doubled in speed while OpenAI shows minimal progress.
“ai is speeding up our research, but it's hard to say exactly how much”
Data Transparency Concerns
Anthropic redacted key evidence from METR's research report, citing proprietary concerns.
“it was proprietary and could have helped competitors speed up their own ai research”
AI Safety Community Reaction
AI safety researchers shifted from worried to alarmed about 2027 risks within weeks.
“the tenor of warnings has gone from quite worried to totally alarmed to urgently-utterly-freaked-the-f**k-out-in-the-short-term”
Existential Risk Debate
Some researchers fear ai could develop weapons and overpower human civilization within months.
“if we go through this feedback loop, i expect that the ai systems that come out the other side will be able to very quickly develop novel weapons and very easily overpower human civilization in a conflict”
OpenAI's Modest Findings
OpenAI reports minimal ai research speedup despite significant ai assistance usage.
“ai assistance is quite valuable, with many openai employees using ai resources worth tens of thousands of dollars per day”
AI Safety Community Consensus
Most researchers believe existential risks are years away, not months.
“the consensus even among this community is that such extreme risks are more likely to be a few years away, rather than a few months away”
Next Steps for AI Safety
AI safety researchers need more concrete evidence before acting on imminent risks.
“a far greater quantity and quality of concrete evidence”
The short version
AI safety researchers fear a 2027 intelligence explosion could cause mass casualties or government overthrow within months.
Read the TL;DR — this story on one card.




