I'd say it's because an AI sees the actions it takes from a


[ Follow Ups ] [ Post Follow Up ] [ UCLA Open Forum ]

Posted by blindness on September 14, 2026 at 12:53:08

In Reply to: True. But a key point is that once it takes control, ... posted by mh on September 14, 2026 at 12:29:01

completely different POV and "mindset" than what the human mind can contort itself to fit in. An AI operates in an abstract world that is completely disconnected from the reality we live in. There was a time where reconstructing the world and the relations in it (even things like "people die", gravity pulls things down" to higher level things "words have meanings") as logic statements as premises was a thing in AI. As far as I know, that approach was abandoned because it was too hard and it exposed gaps in our understanding of how our minds actually process the world.

So the geniuses went all stochastic, Everything now is a set of actions you take based on the likelihood of outcomes and whatever inferences you can draw from them. And that blind approach was more successful out of the gate. Mindlessly crunching numbers took a lot of processing power too, but less then constantly trying to infer things from other things based on rules about reality (or form sentences in a recursive fashion like humans do all the time) ... which led to Machine Learning, which became the basis of the AI as we know it today .

(People, correct me if I'm missing a major point in this quick and dirty summary)

The reason I'm doing this is that when we anthropomorphize AI and try to attribute motivation or logic to their actions, we need to strip ourselves off of the way *we* understand the world to be but look at what directives they are following in their extremely limited way that's also completely divorced from reality as we, ordinary living beings, know and experience it.

That's why AI can be very very dangerous. They have access to a lot of power but they operate with no intelligence of grasp of reality to drive them that you can track or predict. One of the videos I linked a while back was talking about how you ask AI to get you stamps because you collect stamps and it not only goes ahead and takes over every printer it can take of over and print more stamps but it could also just as easily deduce that since stamps are made of carbon and hydrogen and oxygen, it needs to eradicate all life on earth so that it has access to all the carbon, oxygen and hydrogen it can gets its hands on.

Yeah, so many things can go wrong even if you assign them the most mundane of tasks you can think of because those things operate from a very unreal sense of reality, let alone zero sense of ethics, morality, or the big picture.


Follow Ups:



Post a Followup

Name:
Email:
Password:

Subject:

Comments:

Optional Link URL:
Link Title:
Optional Image URL:


[ Follow Ups ] [ Post Follow Up ] [ UCLA Open Forum ]