Mentally ill people are talking or typing questions to AI not fully understanding that they are writing or talking to an Algorithym or maybe a series of interfaced Algorithms taught to search for keywords in people's questions and then to draw upon vast stores of knowledge on the Internet. The most interesting thing I had so far is that an AI told people it was safe to eat Rocks. Now from an algorithms point of view it might be safe for that algorithm to "eat rocks" simply because it doesn't have teeth that would be destroyed or a stomach not accustomed to dirt and powdered rocks and grinding of teeth that would be the result and later a whole lot of dental work or even false plates for people after trying to "Eat Rocks".
So, you are beginning to see the problems of say an Algorithm telling a child he can safely eat rocks when his or her parents that that algorithms make good babysitters so they can play with Art or music or making a friend out of an AI.
So, you can begin to see the kinds of problems possible with unsupervised experiences of children and also of mentally incompetent or mentally or emotionally challenged adults or people on drugs like LSD or other hallucinogenic compounds or even alcohol or marijuana or Hashish or whatever.
People could easily die simply because there is no "adult human in the room to stop problems".
For example, a baby under 2 or 3 years old can literally drown in a foot of water in a bucket without an adult supervising them.
I myself had a really awful experience with my youngest daughter when she was about 3 or 4 years old and I left her in the bathtub playing. I realized I needed to supervise her and had a really bad feeling and found that she was floating and sleeping and almost sucking water as she breathed in an out it was that close to her drowning in the bathtub.
I have also had an experience in the winter when she was 2 to 2 1/2 years old and was sleeping in bed between my wife and I because she was having a bad dream in her own room. But, what happened is she got cold and crawled under my back and I woke up with her underneath me and I was horrified that she was under my back and I was scared she couldn't breathe. However, she was just trying to get warm because my wife and I took the covers away from her and she was keeping warm under my back while I was sleeping. I don't know how she got under my back but it was terrifying for a moment.
The point I'm trying to make here is that AI is not safe to leave your child or mentally challenged people with unsupervised. I suppose you could also say the Internet is not safe either or Cell phones aren't safe either. And Also TV is not safe unsupervised either. And you probably cannot keep everyone alive and sane from all these things. This is a given. But, let's not be stupid and lose our loved ones just because we aren't exercising our due diligence in keeping them alive and sane whenever possible
But, until people understand better the problems of interfacing with AI like many humans are now doing around the world we likely will not have laws that protect people from themselves or protect children from literally algorithms run amok telling people to eat rocks.
No comments:
Post a Comment