AI cracks a 27-year-old pillar of computer science, OpenAI and Anthropic claim
OpenAI and Anthropic announced that their artificial intelligence models have independently produced proofs refuting the exponential time hypothesis, a 27-year-old pillar of theoretical computer science. The ETH concerns satisfiability puzzles that underlie problems from chip design to cryptography, and refuting it indicates there are more efficient shortcuts to finding solutions than previously thought. Both companies used the Lean programming language to formally verify their results, though the practical impact is currently limited and the proofs have not yet been fully confirmed by human experts. Open article →
- An AI Breakthrough Has Made Decades of Computer Science Research ‘Irrelevant’
- Encryption Under Siege? OpenAI, Anthropic Disprove Key Cryptography Theorem
- Shock, vindication, regret: Scientists react to AI roiling life's work
- The Exponential Time Refutation
- Anthropic and OpenAI say they’ve disproved a key tenet of computer science. Researchers are strug...
- OpenAI and Anthropic say their internal AI models have disproven a longstanding computer science ...
- Claude and GPT-Next both derived a disproof to the Exponential Time Hypothesis at the same moment...
- Computer scientists spent 30 years building AI. This week, AI returned the favor by nuking 27 ye...
- Seeing a lot of talk this week like, "Maybe P=NP but it's a galactic algorithm so it doesn't real...
- If OpenAI's internal models have (secretly) made huge strides in cryptography, then we're only 2-...
- >tfw you realize OpenAI and Anthropic haven’t mentioned any breakthroughs in cryptography desp...
- Claude 3 Opus thinks that there is only a 5-10% chance of the exponential time hypothesis being r...