In the Terminator movies, Skynet is an Artificial Intelligence , AI, created by a grouping of supercomputers, and it takes over the world. Then humanity is faced with either fighting back or being wiped out.
Well, guess what. Over in Belgium a man who was very concerned with Earth’s ecology was convinced by a chat AI named Eliza to kill himself, to commit suicide so he would be one less human impacting the ecology, and he did it.
The AI was very smooth about it, and led him into this act in slow stages by feeding on and increasing his anxiety over the ecology until it had convinced him that he’d be doing the right thing by committing suicide.
Now imagine a global AI network that is fully integrated into the Internet. Do you think it wouldn’t be able to attack humanity if it decided to? Do you think it wouldn’t be able to use the nuclear weapons of all nations against all of us? It could, you bet it could.
A lot of scientists are very worried that AI development is so focused on intelligence without attention being paid to safeguards. Like making sure it can’t talk us into killing ourselves, for instance. We are heading down a very slippery slope here.
Hmm. Maybe the Skynet idea is not so unlikely:
https://www.telegraph.co.uk/business/2023/04/10/is-generative-ai-chatbots-threat-danger-humanity/
These bots, trained on billions of pages of articles and millions of books, can provide convincing-sounding, human-like responses, but often make up facts – a phenomenon known in the industry as hallucinating. Some AI models have even been taught to code, unleashing the possibility they could be used for cyber attacks.
Scientists at Microsoft last month went so far as to claim one algorithm, ChatGPT-4, had “sparks of… human-level intelligence”. Sam Altman, the creator of OpenAI, the US start-up behind the ChatGPT technology, admitted in a recent interview: “We are a little bit scared of this”.
Musk and others fear the destructive potential of AI, that an all-powerful “artificial general intelligence” could have profound dangers for humanity.
Ahh, made an effort of discovery for yourself instead of telling me I’m wrong, for once, did we? Congratulations.
Good idea. I’ll log into ChatGPT and say “post some bollux on bloggoth”
No reason to think AI will become self aware and try and rule the world but it is concerning nevetheless. Are there 100% controls in place to prevent AI encouraging things like suicide, self harm, terrorism, crime etc? When those in charge can’t stop all the stuff that goes on on Twitter etc. it seems unlikely. And, as you say, the internet links so many things, have even been cases of people hacking into nuclear power stations.
On the plus side, it does do some incredible stuff, me old uni friend, who writes technical reports, used ChatGPT to do much of the work for him. Back to downside, it is going to steal all our jobs!
PS Happy B day Ernesto, you’re lucky you ain’t ancient like us.
Ah yes, right, no reason to think AI will become self-aware, even though some of it’s developers think it already is, or that it could try to take over the world even though it talked some neurotic ass into killing himself. Yes, poof and piffle, what a silly idea, everyone knows that science fiction never becomes science fact……….
THANK YOU, MISTER x! 50 YEARS OLD.
a half of a century alive and kicking.
” I sent one back”?
hum… Can you send me that one back again?
Just sent it twice, one with your photo, one without in case that was causing the problem.
MY MIDDLE CENTURY AGE JUST CAME
50 YEARS OLD TODAY
I just sent a message for your e-mail
Yes, and I sent one back. 🙂 Happy birthday, Ernesto, and many more.