Isn’t it too easy for the current chatbots/LLMs to lie about everything?

Train it on garbage or in the wrong way, and it will agree on anything you want it to.

I asked DeepSeek about what to visit nearby and to give me some URLs and it hallucinated the URLs and places. Guess it wasn’t trained to know anything about my local area.

I get twitchy by looking at that image, gotta dodge those head crabs.

APAP: All Politicians Are Preposterous?

And then assign an int to a string just to mess with the interpreter.

Breaking news: New Delhi just discovered AGI, she is called Chetana…

Mmm twizzlers…

You are right, do it well-planned and slow.

At least they give it a year, instead of just chainsawing it. That said a year is still very fast, if they did it over 5-10 years that would be right.

The ICE guys face seems to say: If you love your job, then it’s not just a job, it becomes a passion.

The Damascus steel guy Alec Steele, visited not long ago.

They need to keep it running 24/7 because the startup price is much higher than letting it run.

LGTM 👍

How will https://sh.itjust.works/c/curatedtumblr work when Lemmy is Tumblr?

Footage from a robot in skynet mode:

!it’s from the movie Runaway 1984!<

It’s time to sell your tesla stock and invest in cocaine!

Hear me out, what about using JSON to store the configuration in the Python backend?

There will be peace when there’s nobody left to fight.

Cool, good to know.

I wrote a script to do backups on a ESXi it uses Busybox’s ASH, one thing I learned after spending hours debugging my scripts was that ASH does not support arrays so you have to do everything with temporary files.

Still Gandalf was partial to some Old Toby.

Even with the president seat in reach I would expect the GOP to somehow take it from JD if they get the chance.

JD: I’m the president now!
Them: Sure kid…