Subquadratic claims to have unlocked a major LLM bottleneck
2026-06-19 · MIT Tech Review AI · Original
A Miami-based startup has emerged from stealth mode with a bold declaration: they've cracked a decade-long mathematical barrier limiting large language models. While the initial reveal was met with skepticism due to a lack of specifics, Subquadratic is now backing up its claims, offering the technical details to prove they've genuinely advanced AI capabilities.