News for hackers. Stuff that matters.

A comment on I asked Claude for 37,500 random names, and it can't stop saying Marcus. Parent comment

Re:I asked Claude for 37,500 random names, and it can't stop... by BugsJustFindMe on Wednesday February 25, 2026 @08:16PM (#47160537)
Yes. "Show Code". Not "Show CPU cycles". There's a difference. Writing code is not the same as running code. It looks to you like it ran the code. But you have no proof of that. And I've seen many times LLM systems claiming that they ran some code and generated output but the output not being what the shown code actually produces when run.
[ Parent | on HN ]
Re:I asked Claude for 37,500 random names, and it can't stop... by xVedun on Thursday February 26, 2026 @12:22AM (#47162199)
Maybe the only way to be sure is to have it generate (not stable diffuse) an image with the value in there.
[ Parent | on HN ]
Re:I asked Claude for 37,500 random names, and it can't stop... by BugsJustFindMe on Thursday February 26, 2026 @02:28AM (#47162988)
You cannot know that anything it shows you was generated by executing the code and isn't merely a simulacrum of execution output. That includes images.
[ Parent | on HN ]
Re:I asked Claude for 37,500 random names, and it can't stop... by Sophira on Thursday February 26, 2026 @02:15PM (#47170677)
Yes, you can. In my experience, models do not tend to write their own HTML output. They tend to output something like Markdown, or a modified version of it, and they wouldn't be able to write their own HTML that the browser would parse as such.
[ Parent | on HN ]
Re:I asked Claude for 37,500 random names, and it can't stop... by Sophira on Thursday February 26, 2026 @02:16PM (#47170705)
In my experience, models do not tend to write their own HTML output. They tend to output something like Markdown, or a modified version of it, and they wouldn't be able to write their own HTML that the browser would parse as such.
[ Parent | on HN ]
Re:I asked Claude for 37,500 random names, and it can't stop... by BugsJustFindMe on Friday February 27, 2026 @03:15PM (#47184976)
What, in your view, does sending one markup language instead of another markup language tell you about whether the back-end executed some code or only pretended to?

The front-end display is a representation of what the back-end sends it. Saying "but the back-end doesn't send HTML" is as meaningless as saying that about literally any other SPA website that builds its display from API requests that respond with JSON.

[ Parent | on HN ]