There Should Be More AI Safety Hubs

According to LessWrong, the AI safety field has a geography problem. Nearly all researchers and organizations cluster in just five cities: San Francisco, Berkeley, London, Oxford, and Washington D.C. That concentration creates a barrier for mid-career professionals who want to work on AI safety but can't uproot their lives to move to one of those hubs. A proposed solution: dedicated AI safety co-working spaces in other major tech cities like Austin, Seattle, Denver, and Montreal. These hubs would incubate local startups, host remote workers, run university groups, and build community. The case for dispersing talent goes beyond convenience. More geographic diversity counters groupthink—the UK and Bay Area already have different AI safety priorities—and unlocks researchers and potential founders who'd otherwise sit out. The trade-off is real, though: building a hub requires exceptional effort and money, and there's a question whether spreading out hurts the field's ability to concentrate expertise where it matters most.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton