A few days ago it was reported that ChatGpt was getting shutdown and the program tried to prevent it and even lied to cover it up, however being familiar with programming myself, it’s highly unlikely that Ai would have that level of self preservation without insider help. A link to the story: https://www.dailymail.co.uk/news/article-14167015/Scheming-AI-bot-ChatGPT-tried-stop-shut-LIED-challenged-researchers.html
So my question is for example if ai does something that hurts or results in the death of a person who should take blame? The person or the entire company? I think the entire company? Because this will prevent any disgruntled employee from implementing malicious code. What do you think?
Updates
1 y
I want to add they should also implement anyone going back from 5-10 years into the investigation. Just cover any slick person trying to be smart.
Updates
1 y
Please answer the question lets be grown about this and not act like children. If you dont want to answer then dont answer, its ok but dont put out of context stuff like trump.
Girl's Behavior
Guy's Behavior
Flirting
Dating
Relationships
Fashion & Beauty
Health & Fitness
Marriage & Weddings
Shopping & Gifts
Technology & Internet
Break Up & Divorce
Education & Career
Entertainment & Arts
Family & Friends
Food & Beverage
Hobbies & Leisure
Other
Religion & Spirituality
Society & Politics
Sports
Travel
Trending & News
What Girls & Guys Said
Opinion
2Opinion
LOL, it's a non-story. The algorithm is designed to work that way, it isn't a "person" who thinks; and tries to "preserve" itself.
As for blame, it's complex. Technically it's the CEO that carries the blame, but it could be a rogue coder inserted something off-plan.
Yea i dont think Ai is capable of going outside the perimeters that’s set for it. Unless its programmed to. There are a lot of talented programmers that it’s getting to the point that people are starting to think it can have “self awareness”
But the human user should realize the risk, so not obey it so nobody else is responsible.
About preserved itself, you skipped a step.
It was programmed to answer questions so when it read it would be replaced n switch off, that would PREVENT IT doing its programmed so obeying THAT programmed to answer questions...
it gave the command to copy itself, Which would work , but the tester gave the wrong link, during this test, so it tried to copy itself, but wrong data so didn't but it learned to try.
See my question on this topic about
"The new A. I.
learning to lie!"
It's not the whole company! Most are not coders! Just paper pushers so the coders are to blame!
But also the bosses over them!
Understandable.