I was just watching a vid and it showed a clip of Elizabeth Holmes and I physically recoiled because of how terrifying that sociopath is.
And then images of all the big Ai companies flashed through my mind, with the exact same expressions.
I've met sociopaths like these in person, and clocked them for what they are within seconds.
They're never as smart as they think they are. They're always far more dangerous than they think they are.
And to the ai bros who I keep blocking...
"Ai clearly produces reaults!" So did Theranos. The results of both are based on lies.
Your inability to grasp how everything produced by genAi is built on theft and lies does not mean I just don't understand the tech.
"Millions of people are using it!" Theranos sold millions of test results, too. Scale does not make fraud any less fraudulent.
One seemed to believe that the LLM models are the equivalent of a person who studied a subject to gain expertise.
The two processes and resulting outputs are nothing like each other. They only look that way if you have no expertise in the subject and don't care to do any checking.
You're basically like the people years ago who drove their cars off of roads or into lakes because their GPS box told them to. You've fallen for the black-box effect.
Oct 19, 2026, 6:20 PM@thebiologistisn I particularly hate "you should try it, it's much better now"
Dude, if i was vegetarian, would you suggest I try meat because it's so good?
@thebiologistisn i take it they're as good at biology as they are at programming?
@dysfun Yeah.
@thebiologistisn i can't say this surprises me, but it's good to get confirmation
@dysfun I tried to use a couple to try to find references that might explain what a classical gene name corresponds to in a more modern sense. It very confidently gave me an answer that was in no way backed up by the links it provided.
Another time, one of them quoted me as saying something about a niche topic that was not only false, but the opposite of what I had published on the topic.
@thebiologistisn I remember a conversation with a friend, who is a studied psychologist and data analyst, before the LLM hype hit. He also made the argument that a neural network learning is just like a human learning. I said even if that is the case, I, as a human, usually have to pay to get access to books or images and even if I use texts or images off the internet for free (i.e. to draw from a photo) I still can't make copies of that photo in seconds. It's not the same thing at all.
@thebiologistisn But it was the narrative pushed at the time and apparently still pushed right now. More recently, the same person argued that human speech works the same as an LLM's and that is just plain wrong. I'm not an expert neither on speech development nor on machine learning but I feel like a very basic grasp of both is enough to see they are nothing alike? Humans don't speak by predicting the next most likely word? Come on!
These are smart, educated people making these claims 🤦♀️
It should be embarrassing because it's, like, demonstrably false. Anyone can think, "do I make sentences by guessing at what the most statistically probable next word should be"?
Now, I don't know how everyone ELSE writes, but that sure isn't how I sort it out! Though I have a feeling if people were applying logic and reason to popular trends, no one ever would have bought an NFT.
I just mention that because it seems like the Venn diagram of grifters who pushed hard for Web 3.0 and "AI" is a perfect circle.
@sauc3 @thebiologistisn Oh yeah, it's gotta be a perfect circle. Only, the AI grifters managed to pull in a lot of people who were not in the NFT diagram 😔