@EricTopol@NatMachIntell@james_y_zou@suzgunmirac Seems like a v important new benchmark dataset, but the paper tests on legacy models and excludes SOTA models (e.g. GPT 5 pro or Sonnet 4.5). Model updates seem to come out at a pace that exceeds our evaluation + publication pace. How can the field better address this lag?
@rowancheung Little late but created an automation that triggers low power mode when the battery hits 95%. Much more likely to get through the day without recharging
You can just take academic papers and paste them into Gemini 2.5/ChatGPT o3/Claude 4 with the prompt "build me a game based on this paper, make it interesting and thematic but still conveying key findings" and get a tiny working educational game. (In this case, I used Gemini)
I had fun writing this piece and I'd like a lot of people to read it.
I have no idea how to please the algorithm, so I'm asking you, the human on the other side of the screen, to read it, like it, and share it.
https://t.co/lERp4VJB5P