Why I'm Joining Thinking Machines — on democratizing control over AI
ChowdhuryNeil · x · 2026-09-15
Neil Chowdhury published a long-form essay explaining his move to Thinking Machines, arguing that control over AI should be democratized.
- Background: four years of technical safety research at MIT, OpenAI, and Transluce, focused on misalignment
- His thesis: centralized deployment is defended on safety grounds but underestimates the risk of a few companies deciding model values and blocking outside scrutiny
- The tension: releasing weights removes safeguards that assume centralized monitoring; building controls that survive decentralization is one of the most neglected problems in AI safety
- Thinking Machines is betting on AI shaped by its users, and he wants to make that future safe to build
Related event: Neil Chowdhury Joins Thinking Machines to Lead Safety & Alignment(6 posts)→
More from AGI Musings
- Tegmark's "replacement species" AI framing called baseless and dishonest by critic — JosephJacks_ · 2026-10-31
- DeepMind's Economics of Intelligence: AI reshapes economy via agent coordination, not solo model smarts — Scobleizer · 2026-10-31
- Claude can build pro software for ~$20k in tokens — and TAM shrinks to match — bigblueboo · 2026-10-31
- Will AI kill the 9-5? A long read traces the workday back to Josiah Wedgwood's 1769 pottery factory — Konstantine · 2026-10-31
- AI memo: China to restrict exports of 23 AI technologies, plus Kimi K3.1 Pro and Genie 4 — dralandthompson · 2026-10-31
- Chinese firms offer AI 'ghost bots' that resurrect deceased loved ones from 30 seconds of footage — ArtificialOther · 2026-10-31