Sublime

Your team in 2025 trying to A/B‑test a new onboarding tooltip: Cross‑Functional OKR Sync 3 notion pages and a Miro run a pre-mortem "let's push this to q3" Building stealth warplanes in 1976: https://t.co/WvUItmgfHc

Altman is pathetic. First, he says, "We are now confident we know how to build AGI as we have traditionally understood it," then "invents" dumb reminders, dumb "memory" in ChatGPT, dumb search hoping to steal from Google, and now a dumb social network hoping to steal from Musk. https://t.co/2rvtku2bgp

What does DeepSeek R1 & v3 mean for LLM data? Contrary to some lazy takes I’ve seen, DeepSeek R1 was trained on a shit ton of human-generated data—in fact, the DeepSeek models are setting records for the disclosed amount of post-training data for open-source models: - 600,000 reasoning data [1] - 200,000 non-reasoning SFT data [2] - human prefere... See more