Tombmas Spooks
@thomasfuchs.at
18K followers 610 following 75K posts
👨‍💼fine time tracking software at nokotime.com 📸 I killed Flash (with script.aculo.us) 🔭 #astrophotography - https://lightfrom.space 🕹️ #retrocomputing 🐈 #cats Ally. He/him.
Posts Media Videos Starter Packs
thomasfuchs.at
It all started to go downhill when people began to purposefully make vertically video
thomasfuchs.at
This is a skill that should be taught to everyone, I’m asking around if there’s anything helpful (like a website with a game) for this
thomasfuchs.at
Is there a website that trains you to recognize gen AI images and video, perhaps in form of a game?
thomasfuchs.at
Like when it’s a cartoon or TV show it’s very clear that stories are being told, a small child can understand that it’s not showing reality.

But AI stuff mixed in with real videos that are presented like candid camera stuff is different. It’s become good enough to often fool media literate adults.
thomasfuchs.at
Sometimes when something makes you feel uncomfortable your feelings are right. :)
thomasfuchs.at
There’s other stuff like potential cognitive problems as it trains perception to accept impossible things (e.g. items that appear and disappear), especially for children.

Like imagine growing up seeing this stuff for hours each day.
thomasfuchs.at
Besides the obvious deceitfully showing something that never happened as if it was a “candid camera” video, maybe most importantly it will implicitly promote training biases and (usually secret) system prompts.

For example in the “dog jumps on wedding cake” video everyone is thin and white.
thomasfuchs.at
we should have a website for just grinding recipes

nothing could go wrong
thomasfuchs.at
“Why don’t we just use AI to detect AI fakes?”
Reposted by Tombmas Spooks
ignu.bsky.social
i like the wine glass floating in the air
thomasfuchs.at
What about people who like the third season of Star Trek: Picard
Reposted by Tombmas Spooks
thomasfuchs.at
The greatest achievement of “AI” will be the end of the timeline cleanse because we won’t be able to tell if short fun videos are real anymore
thomasfuchs.at
Very impressive hair on this lady
thomasfuchs.at
tbh in this specific case LLMs might actually do a better job than the man
thomasfuchs.at
The super-moon was like almost a week ago, at least fix your writing when you copy and paste stuff for growth hacking your account.
thomasfuchs.at
There's so many people who have the stance that "well, sure AI is bad, but most humans are also bad, so why not just use AI, it's all the same anyway" and this makes mad with a passion
thomasfuchs.at
wtf is this “hype” thing on YouTube and should I even care
Reposted by Tombmas Spooks
thomasfuchs.at
still entertaining offers for a few hours
thomasfuchs.at
It’s amazing how people cannot help themselves but project their humanity into machines.

To wit, a computer doesn’t even know what a number is.

It literally knows nothing.

It’s a machine that runs instructions (that come from humans) to transform one set of bits to another one.
Reposted by Tombmas Spooks
spavel.bsky.social
As an example: AI doesn't understand "no." Because the statements "no ketchup on my burger" and "ketchup on my burger" are almost identical to a machine that does not and cannot actually reason. It's only a 2 letter difference.
AI doesn't know 'no' – and that's a huge problem for medical bots
Many AI models fail to recognise negation words such as “no” and “not”, which means they can’t easily distinguish between medical images labelled as showing a disease and images labelled as not showin...
www.newscientist.com