GPT-5.4 is out. I've been playing with it for the last week, and here's my honest take 👇 What I love: - Speaks more like a human than 5.2/5.3 - 1m token context window - tool use is :chefs-kiss: first model where i've experienced a true "go investigate and fix" experience that feels robust What needs work: - basic front end / UX taste - still loves a bullet point - latency/stability tbd It's likely to move into pole position for gnarly investigations, complex back end work, and things that…