11–16 of 16 posts
I'm not sure "true" reasoning is really necessary. Coding tasks have a lot of guessing what the problem might be and then investigating. Under those circumstances, a hallucination is just a guess that didn't work out.
Earlier quoted context omitted.
You can't predict --- and therein lies the problem. All you can do is verify. And this negates a lot of the value proposition of AI. The best use of current AI is what it was originally designed for --- things that don't matter much and are highly tolerant of errors --- like web search.
Coding agents work well enough to be useful because they can check their own work. Nevertheless it is too generous to call that reasoning. If they're right 80% of the time and they rerun a prompt if the project won't build, they might be right 95% of the time, or even 99% of the time on the third try. And if you know a bit about coding you're probably able to recognize when to intervene. That's not to denigrate a rea…
Which means we’re getting a lot of shitty work that passes tests.
The future of agents is dictated by the future state of LLMs, not the present state.
Earlier quoted context omitted.
You can't predict --- and therein lies the problem. All you can do is verify. And this negates a lot of the value proposition of AI. The best use of current AI is what it was originally designed for --- things that don't matter much and are highly tolerant of errors --- like web search.
[flagged]
> While LLMs are probabilistic, their accuracy in specific domains—like Tool Calling—is already hitting near-100% reliability. That is where industrialization happens.
Is this just an AI bot replying to comments on its own AI post?