Keen Technologies, an artificial general intelligence startup from ex-Oculus CTO John Carmack, raised $20M led by ex-GitHub CEO Nat Friedman and Daniel Gross
Context & Ripple Effects
In August 2022, John Carmack's Keen Technologies raised a $20M round led by Nat Friedman and Daniel Gross — an angel-scale check for an artificial general intelligence lab, struck before the capital race that followed. The related coverage shows how far the baseline moved: Humans&'s $480M seed at a $4.48B valuation and General Intuition's $320M round are both orders of magnitude larger for younger teams.
The backers' own trajectories underline the shift. Friedman and Gross, who also spun up the service that became Andromeda and seeded AIUC through NFDG, have since split into different vehicles — Gross departed Safe Superintelligence, the lab he co-founded with Ilya Sutskever, while Friedman now leads the Meta Superintelligence Labs products team. Keen's round captures the pre-scaling moment when a famous engineer plus two well-known investors was enough to launch an AGI lab.
First-order effects
- Carmack secures roughly $20M of runway to pursue AGI with a deliberately small team, while Friedman and Gross add a flagship founder-bet to a portfolio that already includes the Andromeda service and the AIUC seed.
- The round sets Keen's early valuation and ownership structure entirely on Carmack's track record — ex-Oculus CTO credibility standing in for any product or revenue.
Second-order effects
- Later entrants such as Humans&, backed by Nvidia and Jeff Bezos at a $4.48B valuation, reset buyer and investor expectations so sharply that a $20M AGI raise becomes a rounding error — pushing labs like Keen toward either lean-research discipline or much larger follow-on rounds.
- Friedman and Gross's subsequent moves — Gross leaving Safe Superintelligence, Friedman joining Meta's superintelligence effort — thin out the angel layer that made rounds like this possible, shifting AGI funding toward institutional vehicles and corporate labs.
Third-order effects
- If the pattern holds, AGI labs consolidate into two species: founder-credibility startups whose valuations rest on a named engineer's reputation, and capital-intensive labs requiring nine-figure seeds — with little middle ground, as the gap between Keen's $20M and Humans&'s $480M already shows.
- Repeat-backer networks like NFDG function as the connective tissue of the sector, but as their principals migrate into Meta-scale organizations, the independent angel-funded lab may become structurally harder to sustain.
The trend: AGI startup funding has inflated from eight-figure angel-led bets like Keen's to nine-figure mega-seeds, with the same recurring investors migrating from checks into institutions along the way.