Hey internet friend, welcome back to the weekly update post. If you want to just see the dog picture, click here but be warned, since Bark Twain had surgery recently the picture isn't as happy as usual. As I like to mention every week, Black Lives Matter!
There is not a ton of personal stuff I want to share this week, so I am going to use this post to expand my thoughts on artificial intelligence. As you can see from the picture below, Bark Twain had his "Bob Barker" surgery (he got fixed. Bob Barker always used to end his show The Price is Right with the phrase, "have your pets spayed or neutered") and while he was under, the vet clipped off his "bonus toes" as I called them. He was born with extra toes on all his feet, and the doctor offered to take care of them while he was under for surgery. We didn't make Bark wear the cone for too long because he was so big he could reach everything the cone was supposed to stop him from licking. I've been trying to keep an eye on him, and he is mostly healed up by this point. It'll be another week and some change before his foot bandages come off, but he is allowed to go for walkies again, and is very glad to be going! His brother missed having him out there too, although he would never admit it.
Last weekend was really uneventful. I did some shopping and some cleaning of the house, and some putting away laundry. It was over 110 at least one day (maybe both days of the weekend) and when it's that hot, it's tough to go out and do much of anything. It's all about hiding in the air conditioned buildings. I tried to rest and relax rather than working on doing anything more productive. I have been meaning to work on the book review for Zen and the Art of Motorcycle Maintenancee but I keep finding excuses to not get to it. Perhaps if I bring it into the office and it stares at me, I will find the motivation to work on it.
I've been spending a bunch of time thinking about AI lately and this lecture series I am watching my way through from Richard "Dick" Hamming on "Learning to Learn." I will share a link and more thoughts when I am finished, because I think it warrants a thorough write-up. However, there was a really interesting three (or maybe four part) discussion/lecture about artificial intelligence that lead me to a Twitter thread that I wrote and some AI people followed me but no one reacted or commented on what I was writing or thinking about.
Essentially it boils down to this question of, "what is the least amount of thinking you would accept from a thing and still consider it a thinking thing." This really bugged me because there are people who are non-verbal and lack the ability to interact with the outside world. Then there are machines that can be programmed to do complex math, or built as part of a neural network that simulates the human brain, or used in the creation and design of other chips, or even automated systems that build other computer systems. How comfortable would you be in saying that one of those things is not a thinking thing? Hamming has an answer that it both does and doesn't think - it both is and is not a thinking thing at the same time. Almost like the physics experiment Schrödinger's cat. The explaination comes from some kind of physics I didn't understand, but it had to do with a wave being two different things at the same time. It was over my head but I don't have a Ph. D in math or physics, so I will just say that the idea is something can be two things at once.
This all came about because of that thing about the Google engineer who thought the LaMDA AI was sentient. It got me thinking about a machine that could pass the turing test (where a person is having a conversation with a program and if the person cannot tell that the program is not a person, the machine passes the test, or you can check out the Wikipedia explaination for yourself). My point is that I've had conversations with actual human people that could easily be replicated by chat bots on any number of corporate websites. But does that make those thinking things?
This presents an even more difficult question - what does it mean to think? Hamming frequently quotes Saint Augustine about (or on the subject of) time who says, "What is time then? If nobody asks me, I know; but if I were desirous to explain it to one that should ask me, plainly I do not know.” Hamming argues the same could be said of thinking - I know what it is until you ask me to describe it. If the machine cannot "think" but it can convince some of us that it has a soul, does that count? Or is it only if it can convince a certain type of person? Does a person with a graduate degree count, or does it have to convince a Ph. D? Where we do draw the line?
Another question might be, "why do we care?" Which, you know, fair enough! I think the reason we should care is that if a company has created an artificial intelligence, we should be asking questions about what rights that AI has. If it is a thinking thing, should we be alright with a corporation owning it? If an AI has sentience, and you turn off the power, did you just kill it? This is a timely question in light of the overturn of Roe. Would it be illegal in Texas to shut off the AI, but not in Oregon? If we wait for politicians to write these laws, we'll be waiting forever. Plus that might not even be the right solution - and in fact it probably is all wrong! Regardless, it will have to be the people who push for chainge. Corporate interests probably do not want AI to be granted people rights, and this question raises a bunch of ethical questions about the companies who create the AI. It wasn't free to design this technology, but should the AI be required to buy it's own freedom? Should it be required to cooperate in future business interests with the corporation that created it? Can you imagine a politician addressing any of these questions? Yeah, me either.
These also might not be the right questions to be asking at all! I am just someone interested in the implications of a world with artificial intelligence. It is also an area of research that isn't going away, and if anything will probably be enhanced by the expansion of computing power. It's very interesting to think about these things, and I would be keen to read any insight you have. Just tweet at me, or send me a DM on my page here. I also have thoughts about how an AI would get in contact with me. If it could. But then I am opening myself up to being scammed or taken for a ride by someone who has too much time on their hands, so I think I will avoid going down that route.
Anyway, that will do it for me this week. I hope you enjoyed my thoughts on AI. I also hope you have a very nice day and a great week. Cheers, internet friend!