OpenAI's SQ: "Token use has actually dropped by 2/3" — The New GPT-5.4 'Thinking' Agent Is Shockingly Efficient. - AI Dispatch

OpenAI's SQ: "Token use has actually dropped by 2/3" — The New GPT-5.4 'Thinking' Agent Is Shockingly Efficient. - AI Dispatch | Wave AI Podcast Notes