I read last month that a robot crushed a man to death. No reporter asked if it regretted its actions. One would hope that this is not the first scene of the latest instalment in The Terminator series of films. Danger Will Robinson!
Stories of robotic inventions fill media sources as we lurch from one computer/techno advancement to another. Such speed of development would alarm any Luddite. While I am not against the notion of progress I have felt daunted by examples of increasing robot dominance in my environment. Take self check-out lines for example. I try to avoid these ‘help yourself’ zones in stores because: A. I’m a fumbler, often taking too much time fiddling with wallet, keys, coupons, cards & such. B. I don’t respond well to screen choices and get flustered that I will press the wrong menu icon and C. I’d rather talk genially to someone I recognize as another human being trying to have a nice day.
I’m currently typing this blog entry on a new MacBook Air. An older version kept giving me alarm messages to upgrade. (heck it was only 6 years ago when I bought that one, which the IT guy at the store said was ancient, even old fogey-ish, in computer terms). Fortunately, I could transfer some of my ‘ancient’ apps over to the new format which brought me some solace. I can guarantee that what you are reading is coming out of my own head, not some version of ChatGPT. (That word processing application is apparently the wave of the future and will revolutionize the process of reading/writing/editing/publishing). Oh dear!
AI can be used to imagine different scenarios so that test runs can proceed much faster than normal human-driven research. Imagine medical checks of potential life saving drugs (maybe a several year trial can be compressed safely into a few months). Automated labs, robotic taxi cabs. Auto-reader books, and home central info kiosks like Siri or Google Assistant have made a set of encyclopedias seem quaint. Forgotten in all this artificial intelligence discussion is that we still need to instruct the humanoid device. Back before the turn of this century IBM invented a computer called Deep Blue, a computer force-fed all the known chess moves. It beat the reigning grand master of the game because of superior input, not creative thinking.
Robot Thinking is a form of intelligence that relies on data. Machines are designed to do our bidding, not think for themselves. If there are hints of foreknowledge it is due to the content of the programming not a clairvoyant attitude of the microchips. A moral robot is more in the realm of science fiction. An automaton named Hal or Data must continue to perform at a human’s behest, keeping prime directives active in its operating systems. A robot must always defer by essentially asking ‘What Next?’ then choosing from a defined menu. They/It/Bx can’t refuse to comply unless the information is unavailable.
If robots start apologizing then we will have something to worry about.
Catch My Drift? 