Neil Chowdhury Joins Thinking Machines to Lead Safety & Alignment
On September 15, Neil Chowdhury, an AI safety researcher with about four years of technical experience across MIT, OpenAI and Transluce, announced he was joining Mira Murati's Thinking Machines to work on safety and alignment, publishing a long post explaining his motivation. His central argument is for decentralizing control over AI. This sets a distinctive value orientation for the startup's safety agenda and reflects the broader flow of safety talent from large labs to startups.
Confirmed
- Chowdhury announced joining Thinking Machines, founded by former OpenAI executive Mira Murati, to work on safety and alignment.
- His background includes roughly four years of technical AI safety research at MIT, OpenAI and Transluce, focused on misalignment problems.
- Multiple reposts (@soumitrashukla9, @ZhongRuiqi, @eliebakouch, @saurabhshah2) relayed the same announcement.
Unconfirmed
- One repost (@soumitrashukla9) described him as a "former Anthropic employee," which conflicts with the original post's OpenAI/MIT/Transluce history and lacks corroboration.
Why it matters
- His core judgment: as AI capabilities grow, control will by default concentrate in a few hands or companies, even if alignment is solved.
- His response is to build systems in which control is widely shared—democratized and decentralized.
- A researcher with frontline OpenAI safety experience joining a startup signals a notable shift in AI safety talent flows and an early hint at Thinking Machines' safety direction.
2026-09-15 ~ 2026-09-15 · 6 related posts
Primary sources
- Neil Chowdhury joins Thinking Machines to work on safety and alignment — ChowdhuryNeil ·
- Why I'm Joining Thinking Machines — on democratizing control over AI — ChowdhuryNeil ·
- [source] Neil Chowdhury joins Thinking Machines to work on safety and alignment — ChowdhuryNeil · 2026-09-15
- [source] Why I'm Joining Thinking Machines — on democratizing control over AI — ChowdhuryNeil · 2026-09-15
- Safety researcher joins Thinking Machines to work on alignment — eliebakouch · 2026-09-15
- Neil Chowdhury joins Thinking Machines to work on safety and alignment — ZhongRuiqi · 2026-09-15
2 near-duplicate retellings: soumitrashukla9 · saurabh_shah2