Posted on 02/16/2024 7:04:44 PM PST by algore
A kill switch on that seems an excellent notion.
Exactly
Or throwing the breaker.
The scientists think that the AI will kill them last.
Homo Sapiens are so dumb.
The beliefs of these morons is based upon what they see in movies. Their anxieties can be cured simply by making movies with friendly AI robots.
The race to dominate AI is similar to the race to the ‘Bomb’. Everyone knows there’s huge possibilities and no regulation is going to stop it.
There’s a certain inevitability here. Distributed systems, quantum computing, networks, etc. - with nation states funding ‘anything possible’. If ‘it’ can be done, whatever ‘it’ is, and ‘it’ has the resources, motivation, and means of developing ‘it’ - then ‘it’ will emerge.
There’s certainly a genie trying to escape the bottle - we just have no idea yet what it looks like.
An old friend of mine, an expert in computer and cognitive science, said "AI" was marketing, more or less, for what he called VGP.
Very good programming. Trying to pass the Searle and Turing tests is no measure of intelligence, on the same basis that some people can be convinced by con men and snake oil salesmen. Being convinced is not proof of anything, per se.
The Large Language Models have shown that GIGO remains in operation, and some highly promoted AI experiments ended up adopting racist stances and the like. Making them like dumb homo sapiens.
To this date, we know almost nothing about the most basic cognitive acts, though we homo sapiens -- dumb, as you say -- have been able to mimic them. An informative synonym for mimic is ape, and that's what this AI ( really VGP ) is at this moment in time.
Proof that homo sapiens are dumb. Biden is in office. QED.
A kill switch? That won’t make the AI paranoid against humans at all.
I am still waiting to see anything of remarkable value created by AI.
When AI writes a winning legal appeal, or explains a mystery of human physiology, or indicts a Democrat for election fraud, I will be impressed.
“To this date, we know almost nothing about the most basic cognitive acts”
Great discussion on this podcast:
https://podbay.fm/p/where-is-my-mind
Most of the good stuff is banned and canceled in the university.
The mention of Kurweil and Musk tells a tale, for intertwined are science theory and plain business fact. An element of salesmanship is always present. I suspect we homo sapiens are generations away from scratching the curtains away from this "wizard" problem.
Best wishes and thanks again for the link. I will listen to more while pruning the rose garden this day.
AI is already working on that.
True AI, which we’re still not really near, is self modifying code. So any kill switch has to deal with the concept of how you hide from code that is editing itself the code to kill it.
The best way to solve unsolved problems is to stop banning and censoring people working on them.
Homo sapiens has a long history of failing to meet even that low bar.
The kill switch is a 1970s concept when computers were hard wired, were the size of large rooms and didn’t talk to each other.
It is embarrassing when current scientists talk about it.
AI (in whatever state it is in) will be on most cell phones in a few years.
The adolescence of P-1 by Ryan, Thomas J. (Thomas Joseph),
That's a link from the waybackmachine by the way.
Free inquiry has a bug-a-boo built into it, for all those who would crush what is "free" in favor of "submit" and "obey."
In this, I think world events are "coming to a head," as the "free" contends with the "un-free." I see freedom winning long term.
Best wishes.
AI isn’t self-sustaining. Remove the human technician’s for a little while and any AI system will collapse naturally; this is about controlling people, not machines.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.