AI - 0.10 - intro - random thoughts
A few things to think about before we start:
IBM announced it will "let go" of 30% of its workforce by not hiring new people, to be replaced by genAI.
The companies that are successful with AI are going to be the ones that *increase* their workforce because AI is making their existing employees more productive. If the only way that you can make more money is to fire a bunch of people, and replace them with artificial intelligence, well, I direct you to my thoughts that any friend, counselor, caregiver, or employee who *can* be replaced by artificial intelligence, *should* be replaced by artificial intelligence. The thing is, the companies that are going to succeed are not the ones who replace their existing dull employees with a bunch of dull AI functions. The way that generative artificial intelligence is producing material at present, it is not creative, it is not innovative, and it is not terribly useful. Either artificial intelligence is going to make your existing employees more productive, or you are eventually going to run out of people to fire, and your company is going to go down the tubes anyways.
We constantly forget genAI isn’t human, and assign feelings and intent to the machine.
The only people likely to "fail" the Turing test in this way are those who already treat people like bots. (And, of course, anybody who is so mechanized in their life and work that they *can* be replaced by a machine, *should* be replaced by a machine.)
One of the very strong reasons that I agreed to do this particular series is to try and fight against these perceptions that existing generative artificial intelligence systems have personalities. As we will get into, they do not have understanding, they do not have perception, they do not have emotions, and so trying to relate to artificial intelligence as if it does have emotions is a mistake, and possibly a very dangerous one.
Chinese scientists and engineers are applying ChatGPT-like technology to sex robots, aiming to create interactive, AI-powered companions.
On the flip side of the idea that generative artificial intelligence systems have emotions, is the possibility that we, as human beings, start to relate to artificial intelligence as if it has a personality, and even to prefer to interact with artificial intelligence, rather than with other people. If we are able to create systems and processes that are polite, friendly, patient, and various other attractive traits, and then begin to prefer dealing with our artificial workers, companions, friends, and so forth, we are in danger of losing our ability to deal with the foibles of real people. If we lose that, we lose are actual communities. That is possibly one of the major dangers of dealing with artificial intelligence.
The Tony Blair Institute used ChatGPT to produce a report on the effect of AI on the job market.
This may seem to be amusing, but it points out another dangerous risk. If we start to rely on what are, at present, unreliable systems and helpers, we may start to create material for ourselves, which we come to rely on , and any existing faults or biases that are built into our existing systems, then perpetuate into material upon which we have a greater reliance.
Turing test
In terms of artificial intelligence, Alan Turing is famous for the Turing test. The Turing test says that, when you remove some of the conditions that would normally support our identification of a person, such as their physical presence, then, when communicating through a system that removes the non-text cues, if we cannot determine whether we are interacting with a computer program or a person, then the computer program has passed the Turing test.
Turing may not have been entirely serious when he proposed this test. It may not, in fact, be an actual test which we can use to determine whether we have created something that truly is artificially intelligent. It may be that Turing was pointing out one of the additional fallacies with regard to artificial intelligence, by not defining what we mean by intelligence in the first place. Do we really know what intelligence is, even with respect to ourselves?
AI topic and series
Introduction and ToC: https://fibrecookery.blogspot.com/2026/01/ai-000-intro-table-of-contents.html
No comments:
Post a Comment