I looked into it and the guy has a GitHub for this "program". It basically pre-renders the slop sequences from a set of prompts you input and then lets you choose which branch to follow.

So only this jackass is boiling the oceans to make dozens of videos that will never get seen.

Oh no you have a black fuzzy mold problem too.

I hope yours is easy to clear up, I've tried everything to fix the one in my bathroom but it keeps coming back every night.

Came here to figure out the same thing. It doesn't show up on any of their socials. Sounds like AI garbage start to finish so I think someone tried to start yet another slop band and didn't bother to check if there was an existing artist under that name.

Those are some excellent beans, thanks for sharing them.

How is shipping to the US? I don't want our insanity to be your trouble just because I desire a cat in an adorable hat surrounded by oranges on my wall.

It is. I ran out of characters on an early draft and removed "I agree" from the start of my skeet.

Feels a lot like the first big NFT "auctions" that caught press attention; the winning bids came from the owners of the platform.

smh even the scams are reboots these days.

$30M is a marketing number. The actual amount spent on compute and dev time will be significantly more than that.

I wouldn't be shocked if it comes out afterwards that actual SFX artists had to clean up the output, but by then OpenAI will have gotten their fawning press coverage.

I think you can toot your horn a bit and call pushing 44W through a parallel cable a bit more than "kinda ok". Nicely done!

Do tuxes have a harder time getting adopted than other colorations? I know that their impeccable dapper appearance can be intimidating, but everyone that I've had the pleasure to meet has been a complete doofus in the best of ways.

In the end I got glared at, but I think I deserved it. 🪶

He was apparently intrigued by the strange clicking shadow behind the window. 🪶

The core problem is that an automated mad-lib machine doesn't understand what "truth" IS, but "I can't do that" is undesirable behavior as far as the AI-companies are concerned so they haven't prioritized improving the ability of LLMs to do that.

Users bounce when a system tells them that they can't do a thing, and the primary goal of the ChatGPT interface is to keep users hooked.

It's the same reason Google will completely rewrite your search query instead of returning zero results.

I would love to see your take on a porcupine if that sounds interesting to you.

To be clear this is me with inductance. All of my lab tech experience was dealing with DC signals from measurements with very long time constants.

And the IEEE 1284 standard calls for 62 ± 6 ohms @ 4 to 16 MHz. That's only 45% off, high speed signals are so picky picky!

Meh just tack a 14 ohm resistor inline with each conductor of the data pair.

If you could get your hands on a USB-PD trigger it might be enlightening. Or frustrating! Or both!

I know I would end up spending way too much time trying to get a printer to work over it.