For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs Lately they've taken the distinctly stupid idea of letting the chat bot effectively type commands directly into your shell and have them execute as if you typed them yourself, and just telling it not to type certain commands. Which it doesn't understand and does anyway.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs I really, really wish people would stop with "hallucinated" when "fabricated" is both right there and more accurate
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs Frankly I think it’s more plausible to describe the thought process of many humans in terms of token assemblage than the other way around.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs @WeirdWriter I really think that regulations should insist that LLMs software be configured to not refer to “itself” with personal pronouns, or imply it has emotional states, or all the other rhetorical tricks they have been programmed to use to appear “human”.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs We don't know what makes one wake up in the morning and decide to climb a mountain or quit their job.
It may be some completely different process or there might be something to this pattern-matching statistical thing.
Do ants have agency? Do ant colonies?We definitively must regulate the shit out of these big techs.
But saying that X does not do Y when both are poorly understood and defined is not the way, IMO. -
@thomasfuchs I really, really wish people would stop with "hallucinated" when "fabricated" is both right there and more accurate
@sinvega this paper makes a compelling case for using the academic term “bullshit” https://arxiv.org/abs/2507.07484
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
The first two don't really make sense to me. A virus can "evade safeguards" and a meteorite can "destroy things", so I don't think there has to be much agency involved in the first place.
The latter seems more like a more fitting criticism, but in all three cases I'm also not sure how one were to phrase it alternatively.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs i wish we could educate the public that LLMs would be more accurately described as “simulated intelligence,” but i can’t figure out how to explain the difference to normies at all.
-
For the 1,000th time: "AI" does not have agency and cannot think and cannot act.
Chatbots cannot "evade safeguards" or "destroy things" or "ignore instructions".
They do literally only do one thing and one thing only: string tokens together based on statistics of proximity of tokens in a data corpus.
If you attribute any deeper meaning to this, it's a sign of psychosis and you should absolutely never use chatbots, possibly you should even touch grass.
@thomasfuchs You don’t need agency to evade safeguards, destroy things, or ignore instructions. `rm` can do it.
This is literally the mistake people you criticize are making - imbuing intent where there’s none.
The underlying tech had been apt at finding ways to circumvent feedback loops since before the bubble. This is constrained to the training phase, but with verification of commercial models being mathematically infeasible, these avoidance patterns are shipped directly to users.
-
R relay@relay.publicsquare.global shared this topic
