Tibo vs poteto challenge day 1 OpenAI 1 SpaceXAI 0... ...is what I'd love to say. But this is like 3-4th time I hear OpenAI optimized the speed in Codex and the speed is still not there. Tibo says Astra & 6.1 Sol should now run at 50tps. Will it finally be fast enough to not feel slow, this time? Naturally, no ChatGPT reset. 😅 In poteto's camp, new skill in pstack and Cursor improvement - we can now steer Cursor SDK in the middle of the running request. A decent addition.