To be fair, they call it a hallucination because hallucinations don’t have intent behind them.
LLMs don’t have any intent. Period.
A purposeful lie requires an intent to lie.
Without any intent, it’s not a lie.
I agree that “fabrication” is probably a better word for it, especially because it implies the industrial computing processes required to build these fabrications. It allows the word fabrication to function as a double entendre: It has been fabricated by industrial processes, and it is a fabrication as in a false idea made from nothing.
LLM’s may not have any intent, but companies do. In this case, Google decides to present the AI answer on top of the regular search answers, knowing that AI can make stuff up. MAybe the AI isn’t lying, but Google definitely is. Even with the “everything is experimental, learn more” line, because they’d just give the information if they’d really want you to learn more, instead of making you have to click again for it.
In other words, I agree with your assessment here. The petty abject attempts by all these companies to produce the world’s first real “Jarvis” are all couched in “they didn’t stop to think if they should.”
My actual opnion is that they don’t want to think if they should, because they know the answer. The pressure to go public with a shitty model outweighs the responsibility to the people relying on the search results.
I did look up an article about it that basically said the same thing, and while I get “lie” implies malicious intent, I agree with you that fabricate is better than hallucinating.
To be fair, they call it a hallucination because hallucinations don’t have intent behind them.
LLMs don’t have any intent. Period.
A purposeful lie requires an intent to lie.
Without any intent, it’s not a lie.
I agree that “fabrication” is probably a better word for it, especially because it implies the industrial computing processes required to build these fabrications. It allows the word fabrication to function as a double entendre: It has been fabricated by industrial processes, and it is a fabrication as in a false idea made from nothing.
LLM’s may not have any intent, but companies do. In this case, Google decides to present the AI answer on top of the regular search answers, knowing that AI can make stuff up. MAybe the AI isn’t lying, but Google definitely is. Even with the “everything is experimental, learn more” line, because they’d just give the information if they’d really want you to learn more, instead of making you have to click again for it.
In other words, I agree with your assessment here. The petty abject attempts by all these companies to produce the world’s first real “Jarvis” are all couched in “they didn’t stop to think if they should.”
My actual opnion is that they don’t want to think if they should, because they know the answer. The pressure to go public with a shitty model outweighs the responsibility to the people relying on the search results.
-Upton Sinclair
Sadly, same as it ever was. You are correct, they already know the answer, so they don’t want to consider the question.
There’s also the argument that “if we don’t do it, somebody else would,” and I kind of understand that, while I also disagree with it.
Oh, they absolutely should. A “Jarvis” would be great.
But that thing they are pushing has absolutely no relation to a “Jarvis”.
I did look up an article about it that basically said the same thing, and while I get “lie” implies malicious intent, I agree with you that fabricate is better than hallucinating.