AI hallucinations are not accidental, inevitable phenomena or errors. (lemmy.blahaj.zone)

submitted 6 hours ago by KEI to c/Anti_AI

6 comments fedilink hide all child comments

The deadly power of the word “hallucination”!

Google has not developed real intelligence. Therefore, just as an intelligent living being experiences hallucinations due to confusion in its thought system, artificial intelligence does not experience hallucinations.

An LLM is a language simulation model based on a transformer model and an attention module. An AI hallucination is a phenomenon that occurs due to manipulation of weights.

I’ll explain it simply. The sentence "I was so upset that I bought bread" is a sentence composed of the words [upset], [bread], and [buy].

If the weight is [upset], it outputs [empathy], and if the weight is [bread], it outputs [fact-based]. But if it’s a person, then if a weight is given to [buying], then if a weight is given to [buying], then [synchronization] is output. This synchronization is the essence of AI hallucinations.

Google is using this as a tool to conceal criminal facts.

top 6 comments

sorted by: hot top controversial new old

[-] ThomasWilliams@lemmy.world 1 points 1 hour ago

The weighting is based, ultimately, on the parsing of the sentence derived from training.

For instance, Gemini gave the answer,,"25,000 Ford Rangers were sold in 2025" based on an article which said in response to the question - How many Ford Rangers were sold last year - "25,000 Ford Rangers were sold in 2025 above the level of 2024, making 75,000 in total '.

The website had and cut and paste the answer from a different question - how many more Ford Rangers were sold in 2025 - because when you answer a question the first part of the sentence is usually the answer and the later parts are explanatory.

It is more than likely an AI agent was used to write the website article and this is why hallucinations increase when LLMs feed on their own output.

[-] KEI 1 points 1 hour ago* (last edited 1 hour ago)

The point of my argument is that AI hallucinations are intentional. As you said, the fact that it ultimately occurs through learning means that a person intentionally induced learning. Isn't it? Isn’t creating something that isn’t intelligence, personifying it, and calling it [hallucination] despite the name “artificial intelligence,” a narcissistic idea of British Hassabis, isn’t it?

[-] Peppycito@sh.itjust.works 3 points 5 hours ago

The message is diluted by the infographic. The pretty colours add zero information and instead of thinking about the content I'm left contemplating your design decisions.

[-] KEI 1 points 5 hours ago

So you mean you couldn’t understand? Or are you saying I’m just not good at designing? You mean you don’t want to read it? Are you good at constructing sentences?

[-] Peppycito@sh.itjust.works 1 points 4 hours ago

What does the graphic tell us? What do the colours represent? What do the nesting circles mean? What do the arrows to the words have to do with anything?

I don't have enough information to determine if you're a bad designer, but this one is shit.

Are you good at constructing sentences?

Did I hurt your feelings?

[-] KEI 1 points 4 hours ago

So you want to talk about design issues? Well~ Are you a graphic designer? I wanted to know exactly what you were pointing out. And now I see you. You weren’t interested in the content; you just wanted to criticize it. Thank you for telling me your level.

this post was submitted on 24 Aug 2026

2 points (100.0% liked)

AntiAIAlliance2

41 readers

25 users here now

This community inherits the will of the 'Anti-AI Leader'.

founded 5 days ago

MODERATORS

KEI