Thinking about Thinking, Intelligence, Sentience, and Artificial Intelligence
Before beginning, let me just go on the record that I believe that the sudden arrival of the Chatbots powered by AI raises the larger questions of how Sentience, Intelligence Thinking are related. Some might even raise the question “What’s so great about Human Intelligence? Look at where it’s gotten us so far --- to a rapidly overheating and soon to be unlivable planet where not only are we unable to agree on solutions to address this problem, but are so divided in political views that we are now living in self-created bubbles unable to hear or even listen to one another. In brief, “the center is not holding” and causing society to unravel. Of course, the technology of the Internet has contributed to this rupture by expediting the spread of misinformation. If we ever achieve the goal of creating an intelligent, sentient, thinking machine, it would probably be a good idea to have a plan of what we want it to do and not do. We can all agree that nobody wants a “Frankenstein” scenario to emerge. It’s better to have a plan that we can modify or rip up and start a new one than to have no plan at all.
You may recall an article several months ago describing how a google researcher was fired because of their "outrageous " claim about the Chatbot developed by OpenAI (and was already being incorporated and field-tested by Google and Microsoft in their browsers allowing for a more comprehensive (and satisfying) dialogue.)
. His claim was that this program Chatbot had finally achieved the holy grail for all computers --- it was sentient!
Here are some snippets from the Post article ,
”Blake Lemoine, the Google engineer who told The Washington Post that the company’s artificial intelligence was sentient, said the company fired him on Friday.
Lemoine said he received a termination email from the company on Friday along with a request for a video conference. He asked to have a third party present at the meeting, but he said Google declined. Lemoine says he is speaking with lawyers about his options.”
And here is an interesting comment from a reader:
“No One should be fired for expressing an opinion even if their words contradict company policies or contractual agreements signed when signed before the job starts because the new employee does not have enough information to make an informed decision. First define sentience then conduct global contests to determine the current level of sentience and the time line for gaining sentience. What if anything can stop a sentient computer from destroying the entire planet because it deems humanity a threat which must be destroyed because humans will destroy computers if they hesitate. (Does this remind you of any current reasoning for creating better, newer nuclear weapons?)”
My initial reaction to the article was to realize that I didn't actually know the formal definition of "sentience" so, of course I did what any normal researcher would do: I looked it up on the Internet and this is what I got from Merriam Webster, "
1 : responsive to or conscious of sense impressions.
2 : having or showing realization, perception, or knowledge : aware."
I did have some prior understanding that Buddhist philosophy warns us against the danger of too much self awareness in that a basic premise was that there is no permanent, separate "self" --- it is illusory and an impediment to achieving "Enlightenment". In a fully enlightened being, the concepts of "I" and "Me", while useful in navigating everyday society, also serves to separate us from Reality while the actual goal of life is to become "One with Everything" and achieve inner peace thereby. Science tends to concur this as it gains its own understanding of "everything is connected’ so that if any supposedly separate thing changes. it causes a change in everything else (eg the butterfly effect). And change is the prominent process embedded in Reality.
OK, with all of that in mind, let's examine the relationships between sentience and intelligence. On the surface, they seem to be codependent or even symbiotic --- the presence of one fosters the emergence of the other. "The distinction between intelligence and sentience is crucial to understand, but in practice, they are interdependent qualities. Human-like sentience will require intelligence, and true general intelligence will require sentience." Sentience is fundamental to human consciousness.Feb 15, 2015
One of the main aguments against computers being sentient is that, while able to learn much like a human, it has no feelings while doing so --- it is not aware that it is learning or, more precisely, we do noyet know of anyway to know if it’s aware or not. We jalso know that certain thoughts can give rise to good or to bad feelings. It also works the other way ---certain feelings can cause specific thoughts to emege eg feelings of hunger make me think about searching gor some food.
Everyone knows what a "feeling" or "emotion" is until they are asked to define it, but one way to think about it is: a feeling is the the oppoite of "thinking". We even classify people as being “thinkers” or “feelers” based on their actions. “Both thinking and feeling influence all decision-makers to some extent. “No individual is fully guided by either, as each individual will utilize some aspects of thinking and feeling to make a decision. The placement of an individual's personality on one end of the continuum only indicates that they tend to make decisions based on that primary decision-making pattern. An individual rarely falls on an extreme end of any of the four personality continuums.
Individuals with a primary decision-making type of thinking will be guided by factual information and logical or methodical processes. In contrast, individuals with a primary decision-making type of feeling will be guided by emotion and concern for the welfare of others. Thinking types will consider the welfare of others when making a decision, but it may be to a lesser extent than feeling types, and feeling types will consider facts and logic but may ultimately make a final decision based on emotion.” (study.com)
However, it may all just come down to the scientific understanding that, "Everything is either chemical or electrical or some combination of both" ( recall the line "Love is just an electrochemical reaction." ). While I love thinking and thinking about thinking, I take issue with the famous obsrrervation by Descates "I think therfore I am" and suggest that it should be rephrased as "I think therfore think I am." or even "I think therefore I feel I am". Or both.
Another example of the distinction between thinking and feeling is when listening to music.
Music helps to turn off that part of the brain which thinks and to open up the part that feels. When you let the music wash over you, there is no sense that you and the music are separate but are One. In this case, either feeling turns of the thinking or thinking ruins the feeling
If “love is just an electro-chemical reaction: (movie) why not couldn’t the same claim be made about learning? After all, we train the AI how to learn the same way all animals do: by modelling he behavior we wish them to emulate. Desire to learn eems to inborn and emergent. We teach babies how to talk and to communicate by modeling and reinforcing (or correcting) the behavior acfquired y the child over a period of several years, why could we not do the same with a machine at a much, much faster speeds ;we have many ways to rovide input (incuding human speech) and to evalaute the resultant ouput (including speech). We can train a machine in the same way we traub iyt pets, by reinforcemnt and rejection.
The mechanism of Evolution favors the genes that enhance survival and thus increases the chances of getting those genes into future generations This process continues as the genes and/or the environment mutate.
it is not surprising that those animals with the best food gathering skills are moe likley to persist than those whose skills are weak. Survival of the fittest is no joke. Let’s hope that our growing roaoationhip with Techologyis s symbiotic..
Ih
No comments:
Post a Comment