AI - it's not paranoia if they really are out to get you.

[richard1rst]

Lots of questions on here about the use, pros and cons of AI.

https://www.reviewjournal.com/rj-magazine/in-the-age-of-ai-remember-to-think-like-a-human-2928262/?u...

When Chat GPT-4, the artificial intelligence language-learning model, entered the American consciousness last spring, the internet quickly went on Armageddon watch. To believe the doomsayers, we were witnessing nothing less than a baby Skynet take its first breath.

You know the story. The machines humans make become more powerful, are entrusted with more responsibility and then reach a moment of sentience where they look around, self-aware, and decide human beings are:

A) Idiots. B) Evil. C) Expendable.

… And, more or less, nuke the lot of us.

********

It's only a matter of time before AI discovers Rene Descartes:  I think, therefore I am.

[vintagecraze50]

I do worry about the tech getting into the wrong hands with powerful people who would use it to exploit humans on every way possible. Putin comes to mind.

[vintagecraze50]

But then there was robocop and that machine had a conscientious. But part human.

[vintagecraze50]

It would be the programmer who programs to code to destroy humans. The tech is dangerous in the wrong human hands.

[vintagecraze50]

One more thing. We have enough exploitation going on now. This could multiply it ten times in the wrong hands. I has to be responsibly handled.

[vintagecraze50]

Interesting about the nuke thing. We already have plenty of meglomaniac leaders who are ready to pull the switch.They do not need AI to do it. 

[m60driver]

---

@richard1rst wrote:

It's only a matter of time before AI discovers Rene Descartes:  I think, therefore I am.e

 We are not there yet.  I have been following, from a distance, the development of neural nets since the  1980s when I read McClelland and Rumelhart's book Parallel Distributed Processing while in grad school.  Currently we are in the DOS stage of interacting with large language models.   The one person you need to follow is Geoffrey Hinton, one of the AI "godfathers".  His interviews are both enlightening and frightening.  If you really want to crap in your pants watch some of Alan Thompson's "interviews" with Leta over on his YouTube channel.  And you can blame it all on Canada and those renegades at the University of Toronto.  Resistance will be futile.  

[richard1rst]

---

@vintagecraze50 wrote:

But then there was robocop and that machine had a conscientious. But part human.

---

And in that vein the author touches on both HAL and Data as being opposite ends of the possible outcomes.