StoryFirst seen
Formally verified solutions to problems that stayed open for decades, published at a few hundred dollars per problem, show automated proof arriving as real mathematical work.
Everything it touches
4 names 1 voices 1 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
🚨Google DeepMind’s “AlphaProof Nexus” has solved 9 Erdős problems and the results have now passed peer review: >two if them were open for 56 years >also proves 44 OEIS conjectures >every single step checked by Lean >a few $100 per problem >results are now published in Science all while mathematicians are calling for a boycott of OpenAI after their paper dump. WHAT A TIMING.Read original on X
Google DeepMind's automated math proofs clear peer review in Science
Google DeepMind's automated math proofs clear peer review in Science
DeepMind's automated proofs clear peer review on decades-old math
Google DeepMind's proof system clears peer review on open math
Shared connections
One claimed result would show the Exponential Time Hypothesis is false, a central conjecture in complexity theory.
Businesses can automate more office work when a cheaper model scores the same on real tasks.
Price cuts are pulling large AI spending contracts away from Anthropic and intensifying competition between the labs.
Agents that click and type through software without an API could automate work in tools that lack integrations.