When a model comes out, I usually think about it in terms of my own use. This is largely agentic tooling, and I mostly us Claude Code. All the hallucination and eval talk doesn't really catch me because I feel like I'm getting value of these tools today.
However, this model is not _for_ me in the same way models normally are. This is for the 800m or whatever people that open up chatgpt every day and type stuff in. All of them have been stuck on GPT-4o unbeknwst to them. They had no idea SOTA was far beyond that. They probably dont even know that there is a "model" at all. But for all these people, they just got a MAJOR upgrade. It will probably feel like turning the lights on for these people, who have been using a subpar model for the past year.
That said I'm also giving GPT-5 a run in Codex and it's doing a pretty good job!
I’m curious what this means. Maybe I’m stupid but I read through the sample gpt-4 vs got-5 and I largely couldn’t tell the difference and sometimes preferred the gpt-4 answer. But like what are the average 800 million people using this for that the average 800 million user will be able to see a difference?
Maybe I’m a far below average user? But I can’t tell the difference between models in causal use.
Unless you’re talking performance, apparently gpt-5 is much faster.
4o would start writing immediately without thinking. So if the first thing it wrote was “The world is flat because…” then it will continue to write as if the world is flat.
It makes it very stupid, but very compliant. If you’re mentally ill it will go along with whatever delusions you have, without any objection.
https://www.interconnects.ai/p/gpt-5-and-bending-the-arc-of-...
When a model comes out, I usually think about it in terms of my own use. This is largely agentic tooling, and I mostly us Claude Code. All the hallucination and eval talk doesn't really catch me because I feel like I'm getting value of these tools today.
However, this model is not _for_ me in the same way models normally are. This is for the 800m or whatever people that open up chatgpt every day and type stuff in. All of them have been stuck on GPT-4o unbeknwst to them. They had no idea SOTA was far beyond that. They probably dont even know that there is a "model" at all. But for all these people, they just got a MAJOR upgrade. It will probably feel like turning the lights on for these people, who have been using a subpar model for the past year.
That said I'm also giving GPT-5 a run in Codex and it's doing a pretty good job!