samir

Discovery Loop applies AI to run millions of parallel scientific experiments (materials for fusion, better solar cells, batteries) with rapid validation cycles, unlike other neo-labs that lack a clear path to differentiable value capture.

Fund strategy: bold early impactful investments with technology edge, no market risk, collaborative 5-day-office culture, no deal attribution; pro-rata only when max allocation or to support deserving follow-on, otherwise 3x or 1/3x pro-rata.

Discovery Loop, founded by ex-Google leaders including Jeff Dean, focuses on recursive self-improvement (RSI) and world models to automate scientific experimentation, a market potentially as large as frontier LLMs ($100B revenue, $2T market cap), with private company advantage in focus and risk-taking.

Discovery Loop, founded by ex-Google Chief Scientist Jeff Dean and team (80 collective years at Google), is pursuing Recursive Self-Improvement (RSI) for scientific discovery — designing thousands of parallel experiments in closed loops; Samir's fund co-led the round believing RSI could be as profound as frontier models ($100B revenue/$2T market cap potential), with private structure allowing more risk-taking than public labs.

Samir's fund was the only venture investor in OpenAI's first round and maintains conviction that frontier model leaders like OpenAI and Anthropic could reach $100B revenue and $2T market cap, validating the scale of opportunity for RSI ventures.

Samir's firm co-led the first round for Discovery Loop, founded by Jeff Dean (ex-Google Chief Scientist) and team, betting that recursive self-improvement (RSI) and world models will automate scientific experimentation — a $2T market cap opportunity if RSI matches frontier model impact.

Discovery Loop, founded by former Google chief scientist Jeff Dean, aims to use world models and recursive self-improvement to design thousands of parallel experiments for materials science and battery technology, a frontier as profound as LLMs with potential for $100B revenue and $2T market cap.

Coastal Ventures incubated Commonwealth Fusion from early stages, representing their deep-tech/hard-tech investment model — long time horizons, technical edge, company-building support rather than financial engineering.

Exemplifies Coastal's early-ownership model: first investor in a New Zealand rocket company nobody watched, held 28% through IPO, now a multi-billion dollar public company. Validates their 'bold, early, impactful' thesis over party rounds.

Discovery Loop applies AI to run millions of parallel scientific experiments for materials discovery (fusion magnets, solar cells, batteries), with rapid feedback loops similar to coding agents. Jeff Dean's 27-year Google legacy and team departure signal exceptional conviction.