Erdosgate
We recorded 6 claims from this post. 3 were checked against evidence. 3 can’t be checked. 1 more is held. Read the post at its source.
Claims from this post (6)
In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.
Assessment: SupportedSaid
“Alas, “found a solution” didn’t mean what people thought it did.”
What we tested: Bubeck's claim that GPT-5 'found a solution' to Erdős problems did not mean original solutions as people understood it.
What happened: The database maintainer confirms that the findings were references to existing solutions. His accompanying explanation distinguishes his database's open label from proof that no solution existed. (Thomas Bloom, X, )
Assessment: SupportedSaid
“All that really happened is that GPT-5 (oversimplifying slightly) crawled the web for solutions to already-solved problems.”
What we tested: GPT-5 did not solve open Erdős problems; it essentially located existing solutions to already-solved problems on the web.
What happened: Bubeck directly characterizes the episode as work with existing knowledge. His detailed example identifies an earlier published solution rather than a newly proved result. (Sébastien Bubeck, X, )
Assessment: Can’t be checkedSaid
“I smell a really big dose of people believing what they want to believe.”
What we tested: Quick acceptance of Bubeck's claim was likely driven by people believing what they wanted to believe.
Assessment: Can’t be checkedSaid
“The claim would have been extraordinary, and should have been vetted closely.”
What we tested: Bubeck's Erdős claim, which would have been extraordinary, should have been closely vetted.
Assessment: UnresolvedSaid
“I found that the claim of “solution” was pretty misleading, as I summed up in a tweet, and they refused to correct their misleading presentation:”
What we tested: In 2019 OpenAI's claim that its robot 'solved' the Rubik's cube was misleading, and OpenAI refused to correct it.
Assessment: Can’t be checkedSaid
“Some things never change.”
What we tested: 'Some things never change': a closing quip linking the Erdős episode to OpenAI's 2019 Rubik's cube 'solved' claim, which Marcus found misleading.
Held from this post (1)
These records did not pass every check, so nothing about them is published. Held is not a finding.
- The criteria are ambiguous. Our reading, not assessed: The oversold Sparks of AGI paper dubiously alleged that GPT-4 could reasonably be viewed as an early, yet still incomplete, version of an AGI system.