An interview with Skype co-founder Jaan Tallinn on investing $100M+ in 100+ AI startups and why he thinks he failed to steer AI development toward human safety
not because he was excited about the technology — but because he believed it was a threat and hoped he could steer it in a safe direction. Now, he's admitting defeat. https://www.semafor.com/... Ben Smith / @semaforben : Another glimpse into the pre-history of AI from @ReedAlbergotti https://www.semafor.com/... @semafor : “My philosophy has been that I want to displace money that doesn't care,” he told @ReedAlbergotti. “Plan A failed. There is a dissonance between privately being concerned and then publicly trying to avoid any steps that would address the issue.” https://www.semafor.com/...
Context & Ripple Effects
Jaan Tallinn's admission closes a loop that opened with his own portfolio: over $100M across more than 100 AI startups, deployed not out of enthusiasm but as an attempt to buy influence over how AI develops — a strategy he now calls failed 'Plan A.' In the same Semafor conversation cycle, Vinod Khosla laid out the opposite thesis, framing AI as liberation rather than threat.
The confession also foreshadows a broader unraveling of the safety-via-capital doctrine: months later Tallinn went further, questioning whether companies should be run on effective altruism principles at all, while Max Tegmark of the Future of Life Institute kept the existential-risk argument alive outside the investment channel entirely.
First-order effects
- Tallinn publicly concedes that privately worrying about AI while funding its builders produced 'a dissonance between privately being concerned and then publicly trying to avoid any steps that would address the issue' — stripping moral cover from his own $100M+ startup portfolio.
- The 'Plan A failed' framing hands Tallinn's co-investors and the safety-aligned founders he backed a credibility problem: their capital can no longer be pitched as a steering mechanism.
Second-order effects
- Accelerationist voices like Khosla gain narrative ground against doom-adjacent investors, since the most prominent attempt to steer AI through equity stakes has just self-reported failure.
- Other funders who justified AI positions as safety leverage face the same reckoning, pushing safety advocacy toward non-market channels — the route Tegmark and the Future of Life Institute already occupy.
Third-order effects
- If investor-side steering is conceded as dead, AI-safety influence migrates from portfolio construction toward regulation and public pressure, splitting the field between state interventionists and the capital-acceleration camp.
- Tallinn's own drift away from effective altruism suggests the movement's core mechanism — deploying wealth to steer technology — may not survive contact with frontier AI economics.
The trend: AI-safety capital is abandoning the invest-and-steer model, with its most prominent practitioner declaring it failed and the debate re-polarizing around accelerationists like Khosla and external-pressure advocates like Tegmark.