Originally Posted by
Iwokeup thisway
My son has a MSEE and has been relating some scary tales to me regarding AI.
Job interviews for a remote job at his company that were held via video-conference, have been postponed in favor of in-person. (I’m just a 63 year-old dumb seat jockey so forgive my ignorance). AI images of a candidate showed up to the video interview. He said it was extremely difficult for HR and the interviewer to distinguish from real, the AI (chatbot?) gave the answers to the interviewer’s questions. My son’s team found out somehow after the “person” was hired that the work was actually being performed by an overseas company.
A similar situation occurred when an actual worker showed up in person for training after being hired. The guy who interviewed the worker didn’t recognize her. The interviewer later reviewed the video interview and confirmed it wasn’t the same woman. When confronted, the worker stated she was the voice behind the AI generated image appearing on the video. Apparently she wanted to enhance her looks in order to get the job.
Another example my son cited were newly hired graduates who use AI almost exclusively for problem solving and coding programs. They put so much reliance upon the AI results that few bother to validate the AI responses. After several expensive errors, a new job category of AI validators had to be created to check the answers.
Sorry for the long answer, but I was completely unaware that AI was so prevalent in society. Eventually it will get figured out, but I think I’ll be long retired before HAL will be sitting in my seat.
AI works off of incomprehensible levels of information/data.
This amount of information that now can be analyzed, such that patterns are detectable, can be the source of levels of power and control that until now was fantasy.
From a side comment in an essay on globalization:
"the job losses and restructuring that will come from AI will make what happened from globalization a mere footnote"
AI, while not conscious, is intelligent. It can, and already does, make independent decisions. Some of it is smarter than us by many, many magnitudes.
I have read that even the deep dive computer theorists are unable to explain exactly how advanced AI works. (kind of like a fuel control unit /s)
Short of powering down the world's electric grid (which AI would attempt to thwart), advanced AI cannot be "unplugged". Like people networks, it is now everywhere,in computer networks/clouds etc.
Quantum computers are coming.
Can you imagine AI and Quantums's children?
Perhaps we will find that higher intelligence is an evolutionary dead-end.
(ala Bostrom's "maximize paper clip production")
The different paths AI can lead us on are almost infinite.
It will take both wisdom and political pragmatism to choose somewhat safe roads.