site banner

Friday Fun Thread for January 9, 2026

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

2
Jump in the discussion.

No email address required.

Why are advertisements for AI so bad?

There's one running during football games where a coach is, I think, supposed to be picking players for the draft. And he starts asking the AI to give him the linebackers with various traits, then asks for the ones with "strong leadership abilities." And that is OBVIOUSLY A TERRIBLE QUESTION to ask AI! All an LLM could maybe do is search news articles to see if any have been called out for it, but in all honesty I'd expect it to tell me "yeah no can do boss." Like there's probably a lot of useful things an LLM can do for NFL draft prep, but asking it to assess intangibles is not one of them!

And there have been others just the same. Apple ran a series of ads where employees used AI to just not do their jobs. Like a producer using it to summarize a script that she then agrees to buy (the benefit being that she didn't get caught not reading it). Or using it to pretend to participate in a meeting you aren't prepared for, or reply to emails. And the impression I get out of it is that as an employer I would not want my employees using AI to make it harder to see if they're doing their jobs.

It just seems like they're giving terrible examples of awful and irresponsible ways to use LLMs which will almost certainly lead to disappointment and disillusionment.

My experience has been that AI is really good for personal life stuff, but I almost never use it at work. I'm sure that eventually firms will learn how to use it effectively*, but right now it's a helpful consumer product and less of a workplace tool.

*And I understand that it's basically there for coding already

And I understand that it's basically there for coding already

So far that hasn't been my experience. It gets uncannily close,and if you don't pay attention it's fine, but for any reasonably sized project the code tends to be rife with bugs. It's not just one kind of bug either - it varies from braindead stuff like not closing resources to subtle misunderstanding of APIs and business requirements.

The consistent thing I see from its biggest cheerleaders in my company is a kind of Gell-Mann amnesia, where they use it to do something they themselves don't know how to do, and are then blown away by the "quality" of the output. I end up getting assigned to fix it a few weeks later after the hype has died down.

And I understand that it's basically there for coding already

That is very much a subject of controversy in the industry. Some people say that is true, others (myself included based on my experience) say that it ultimately slows you down rather than speeds you up.