AI Math vs. Human Math & Tariffs in the Midterms 10/8/26

OpenAI

Across the shows: 18 speakers in other episodes, 10 bullish and 18 bearish calls, past 30 days

• Tristan Buckmaster · New York University (Courant Institute)Bearish

  OpenAI's user-data protections are weak, given a Claude-assisted hack of its internal systems.

  Buckmaster says a model-training setting was left on by default on one of his personal accounts, which he did not switch off. He argues there are not the walls people believe between his work and user data at OpenAI. He points to a white hacker who used Claude to get into OpenAI's internals and access user data, and says OpenAI's sandboxes are not holding up at all. He suspects his work was collected from his account.

  “I mean, you know, just recently, they had a white hacker who used Claude to hack into their system and gain access to their internals and get access to user data.”at 31:19

• Tristan Buckmaster · New York University (Courant Institute)Bearish

  OpenAI reached his unpublished Navier-Stokes result within a day, using his methods after learning his name.

  Buckmaster says that before OpenAI learned his name it had no progress on the problem. After it did, its agents started from where his team left off. He says OpenAI used about 100 agents, or 10,000 by another account, and got to exactly where his team was, with the same argument and methods, within a day. He notes OpenAI's account has shifted, including its date for the private data used, and that it has not denied the tactics he described on the call.

  “They used a 100 agents and they got to exactly where we were with exactly the same argument and exactly the same methods within a day.”at 33:26

• Tristan Buckmaster · New York University (Courant Institute)Bearish

  OpenAI's recent paper release wiped out early-career mathematicians' research programs.

  Buckmaster says OpenAI's batch of papers released on Tuesday destroyed the careers of early-career mathematicians. He says whole research programs were wiped out by that release, so the speaker sees AI output from OpenAI as displacing human mathematical research work.

  “the recent dump of papers on Tuesday which destroyed the careers of early career mathematicians. So we've had whole research programs wiped out on Tuesday by their dump”at 36:51

• Tristan Buckmaster · New York University (Courant Institute)Bearish

  OpenAI has a massive conflict of interest: its panelists see researchers' private proposals and can front-run customers.

  Buckmaster argues OpenAI has a conflict of interest because research proposals are reviewed by panelists whose summaries and knowledge can flow into models. He says that for the next few years OpenAI will have all the mathematicians' and scientists' research programs in its training data and can see how they plan to proceed. He reads this as disdain for customers, since a company that can look at a researcher's private notes and beat them to market gives customers little reason to buy its product.

  “they have a massive conflict of interest right? They have all they for the next few years they have all all the mathematicians and scientists as research programs in their training data... the fact that they would front run their customers. I mean it shows their sort of disdain for their customers.”at 37:26