Babuschkin's River AI Raises $1.1B to Rebuild the AI Stack Around Personally Trained Agents

Skynet Chance (+0.03%): Democratizing reinforcement learning and fine-tuning of open models to anyone with an API key diffuses capability-shaping tools beyond centralized labs, widening the surface for misaligned agents even as it reduces single-actor concentration risk. The net effect is a modest increase in loss-of-control probability through proliferation.

Skynet Date (+0 days): A billion dollars of capital plus Nvidia and AMD backing accelerates buildout of cheap, fast post-training infrastructure, shortening the interval before highly capable autonomous agents are ubiquitous.

AGI Progress (+0.01%): The claimed large reduction in RL fine-tuning time and cost, if real, lowers a significant bottleneck in post-training, currently a primary driver of frontier capability gains. It is infrastructure and funding rather than a demonstrated capability breakthrough, so the impact is moderate.

AGI Date (+0 days): Massive early-stage funding, chipmaker participation, and an explicit plan to rebuild training and hardware end to end add another well-resourced competitor pushing the frontier faster.