Peter Thiel knows AI might kill us all. He still says accelerate. In a 65-minute conversation with Axel Springer's CEO, Thiel lays out why. The argument is more rigorous than it sounds: 1. The stagnation thesis Decades of technological slowdown have already been killing us — slowly, economically, politically. AI is the first genuine break in that pattern. 2. Slowing AI loses even at 10% extinction risk If Western democracies pump the brakes and China doesn't, you haven't reduced risk — you've just handed the outcome to a different actor. The math on deceleration only works if it's global. It won't be. 3. China weaponizes Western AI fear Every time the West debates pausing, Beijing reads it as an opening. Western safety concerns are being converted into a strategic advantage for an adversary with no equivalent debate. 4. Europe is already behind — and falling Thiel compares Europe's position to Minitel — the French internet that technically worked but was lapped by history. The regulatory instinct that feels like caution is actually surrender. 5. The political instability feedback loop Stagnation produces Weimar-style politics. Thiel gives a democratic socialist 50% odds of winning the White House in 2028 — not because he wants that, but because stagnant economies radicalize electorates. The core bet: the risk of moving fast is real but bounded. The risk of stagnation is already happening, and its political consequences may be unrecoverable. That's not recklessness. It's a different risk calculus than most people are running.