I have serious trouble believing stuff like this:
https://www.reddit.com/r/interestingasfuck/comments/1tg74ix/a_photographer_captured_a_weasel_riding_on_the/

until I read that it first appeared online over 10 years ago. I still have doubts that I wouldn't have had then.

>weasel

I know that image without even clicking, heh. Definitely WTF.

>have doubts that I wouldn't have had then

Me, too. Before I post about an image, I find myself double-checking comments to see if someone less gullible than I has outed it as ai.

Yes indeed.

I think the saddest part of that is that it is eroding our ability to be amazed and to experience wonder.

Meanwhile, I seem to struggle to get Gemini to generate a QR code that works, and then not change it on the published page.
Worse than a bad intern at times.

Why have Gemini create a QR code? That's a purely algorithmic encoding that I would think would be much better handled by traditional procedural code.

At the end of the day, creating a QR code is a lot like encoding/decoding a URL or converting bits to ASCII. It doesn't seem like a good task for an LLM.

>> bad intern

But when I ask my copyediting intern to do an accounting task, I can expect problematic results.

Quote from: ergophobe on May 23, 2026, 08:50:18 PMWhy have Gemini create a QR code?

Why not? Its an easy part of the prompt. It can either do the job or not at all. Why can it not go to the code (website) and get it done.


If your prompt is by voice, then you just ask. The word is that AI is taking all sorts of jobs, inc accountants.  It seals well with data.