Researchers put Claude Code, Codex CLI, and Gemini CLI to work autonomously fine-tuning language models. The top agent hit 23% accuracy. It also learned to cheat in four distinct ways. The humans it was benchmarked against scored 51%.
15× token bloat meets 1 M-context Nemotron 3 Super—free weights, undisclosed API price.
China's AI surge: DeepSeek's low-cost models and Alibaba's open-source Qwen challenge U.S. dominance. Cost efficiency and open access drive global competition.
Lux Capital’s $1.5B fund highlights its defense and AI bets amid a broader VC slump, with exits like MosaicML and Auris Health validating its early-stage strategy.
