211–220 of 281 posts
Excellent. Yes, please give free users more access and leave paying codex users in the dust. Wise plan, Sam.
Earlier quoted context omitted.
How are the models too dumb? How are they dumber than the average person? I wonder if anyone gave Claude an IQ test (the one for humans).
I just asked Fable 5 max to create a 2d game about caterpillar climbing a tree and eating fruits. The game looks good - animations, 8-bit aesthetics, procedural tree branching, but the tree's branches are dead ends. You can't go back once you started climbing a branch. Yes, LLM doesn't have a reliable way to test it's game yet. All the screenshots, and playwright tests will never be enough to test even a simple game.…
For me on a paid plan, the effort indicator was hidden and the model was 5.5 instant. I had to press the + button to select “think harder” before the dial that allowed Sol medium or high to be selected to show. It made me wonder how many paid subscribers realize they are using the same 5.5 instant model as free users by default. A dark pattern or oversight?
I’ve literally seen people ask ChatGPT for a link to Gmail.
Is it a dark pattern, or design decision that makes users happier?
This post makes it sound like paid users should be getting a 5.6 Nonthinking as the Instant option. Still looks like 5.5 Instant to me?
>avoid extra detail when it does not help I wonder if they actually do it to optimize inference. I maintain a corporate AI server and one of the tricks to reduce the load was to modify the system prompt to be as terse as possible so the average response completes faster and requests queue up less often.
Less verbose output = less context & less token gen.
This is actually a downgrade for free users since currently it uses GPT-5.5 for a few messages before it drops you down to GPT-5.5-mini. Now it always uses a model worse than Mini (Luna is nano-equivalent, "It roughly corresponds to the nano model tier used in earlier GPT-5 families." https://developers.openai.com/api/docs/models/gpt-5.6-luna ). I guess it's a bit better with Thinking though. They should use Terra fo…
Excellent. Yes, please give free users more access and leave paying codex users in the dust. Wise plan, Sam.
I dunno, I've honestly never been happier as a Codex customer. Sure, there's been less bonus usage resets recently, but the amount of productivity and value I've gotten out of my $20/month personal subscription is bonkers.
If you're price insensitive, Fable 5 is definitely still the best, but a lot of people aren't.
Earlier quoted context omitted.
If we could show the current models to someone like Alan Turing, I am sure he would conclude that we have AGI.
You say that they pass the Turing test yet every post on HN complains about the way LLMs write so clearly they haven’t passed it yet because we can still tell it’s a bot.
Earlier quoted context omitted.
I wonder if it’s a version of Dunning-Kruger effect to call AI models dumb. I haven’t seen a “dumber than me” model since years. Also the smartest people known in the world use them in their fields so I don’t know what is meant by a “too dumb” model.
You need to be smarter (or rather: more knowledgeable in the problem domain) than the model to be able to use it efficiently. Hallucinations are still a problem occasionally but a bigger one is failure of imagination. Even Claude Fable lacks a holistic understanding of many domains it wasn't obviously trained on. The biggest problem with AI (if we assert that LLMs can be the basis of AI) is that these models will mak…
This sounds interesting on it's own. I would be curious to hear more if you are willing to share.
Earlier quoted context omitted.
And after half an hour using it he'd just admit that his test was way too simple as these models are still way too dumb
How are the models too dumb? How are they dumber than the average person? I wonder if anyone gave Claude an IQ test (the one for humans).
Just look at some training sets to see how the sausage is made: https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-...