Ex-AI Security Institute researcher joins Geoffrey Irving's new alignment lab Resolution
geoffreyirving · x · 2026-10-20
Researcher Siddarth V announces a switch to technical alignment research on RL, joining the scalable oversight team at Alignment Science Research (resolutionorg) in Berkeley — a new alignment bet by Geoffrey Irving and others from the AI Security Institute and Timaeus Research.
His stated reasons:
- We are not on track to solve superintelligence alignment;
- Some version of recursive self-improvement (and soon superintelligence) is likely near;
- More capabilities researchers should switch to alignment.
Irving welcomed him, predicting a coming wave of ex-capabilities researchers moving into alignment and safety.
More from AGI Musings
- AI filmmakers aren't replacing Hollywood — they're starting a new indie film movement — DavidmComfort · 2026-10-20
- Alignment researcher Alex Turner: shutting down rogue AI agents took over a week — Turn_Trout · 2026-10-20
- Midjourney CEO has an AI swarm rate every AI lab on YOLO vs FAFO vs FOMO — DavidSHolz · 2026-10-20
- Researcher proposes free 2-hour access to top AI models before elections to aid voters — NathanpmYoung · 2026-10-20
- World Mechanics Co-founder Pushes Back on Self-Improving LLMs Solving Physical AI — soniajoseph_ · 2026-10-20
- Law firms license every AI tool but never learn which fits which work — jkubicki · 2026-10-20