GPT-5.6 Sol cracks 30-year statistics conjecture via novel synthesis
OpenAI's GPT-5.6 Sol Pro solved a three-decade-old statistics conjecture about the Benjamini-Hochberg method in 90 minutes, where GPT-5.5 failed after 20 hours of computation. The solution recombines established techniques rather than introducing wholly novel mathematics, raising a critical question for the field: whether frontier models generate original insights or excel at synthesizing existing knowledge at scale. This outcome matters for researchers evaluating AI's role in mathematical discovery and for understanding the boundaries between pattern completion and genuine innovation.
Modelwire context
ExplainerThe detail worth sitting with is that GPT-5.5 failed after 20 hours while GPT-5.6 Sol Pro succeeded in 90 minutes, which means this is as much a story about a discrete capability jump between adjacent model generations as it is about AI and mathematics broadly. That gap is unusually sharp and the mechanism behind it has not been publicly explained by OpenAI.
This story sits largely disconnected from recent activity in our archive. The closest thematic neighbor is the Suno training-data breach covered on July 15, which raised questions about what AI systems are actually built on and whether that sourcing is disclosed. That framing applies here too: if GPT-5.6 Sol Pro is recombining established mathematical techniques rather than deriving new ones, the quality and breadth of its mathematical training corpus becomes a material question, and one OpenAI has not answered publicly. The Suno story shows that training provenance tends to stay opaque until something forces it into the open.
Watch whether an independent mathematician or a University of Pennsylvania collaborator publishes a formal verification of the proof within the next 60 days. Peer-reviewed confirmation would separate a genuine result from a plausible-looking but flawed synthesis.
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsOpenAI · GPT-5.6 Sol Pro · GPT-5.5 · University of Pennsylvania · Benjamini-Hochberg method
Modelwire Editorial
This synthesis and analysis was prepared by the Modelwire editorial team. We use advanced language models to read, ground, and connect the day’s most significant AI developments, providing original strategic context that helps practitioners and leaders stay ahead of the frontier.
Modelwire summarizes, we don’t republish. The Decoder originally reported this story as “GPT-5.6 Sol reportedly disproves a 30-year-old statistics conjecture in 90 minutes after humans couldn't crack it”. The full content lives on the-decoder.com. If you’re a publisher and want a different summarization policy for your work, see our takedown page.