Over the past 6-12 months, I've had a number of conversations with people regarding "artificial intelligence" (AI). Most people are afraid of it, for the same reasons they fear anything - lack of knowledge of a particular subject. Rather than educate themselves on the facts, most people go with whatever media, friends, neighbors, etc. tell them.
"X" is bad! It's evil! It's unnatural!
More often than not, a little bit of education leads to the realization that "X" is not bad/evil/unnatural/scary. History shows that once upon a time women were property, people of African descent were only 1/3 of a human, homosexuals are out to get your kids, etc. The facts show that none of this is true. Yet people tend to listen to gossip quicker than actual reality.
But what about this 'AI'? It's man-made. Corporations are fighting for control of it. Of course they are! It has amazing profit potential for capitalists. Remember, corporations usually exploit rather than originate, then market a bunch of BS about it, and make as much money as they can from it. They 'capitalize' on it. AI is just the latest 'big thing'.
What is AI? Is it really coming to take over? Simply, no.
![]() |
Jack D. Myers MD |
I've known about AI since I was a kid. Our dad was a cardiologist. A friend and former instructor of his, Jack Myers, was a genius in the world of medicine, especially diagnostics. In the 1970s, he, along with computer scientist Harry Pople, developed INTERNIST-1 - a computer-aided medical diagnostics system. The system was based on his vast clinical expertise in diagnostics. This later evolved into the "Quick Medical Reference" (QMR) system still used in some medical schools and hospitals. If you're not sure what I mean by diagnostics, go watch an episode or two of the TV show HOUSE. Myers and Pople were the origins of AI in medicine. Did their system put doctors out of business? Again, no. But I bet you didn't know this existed! In our family, this AI has been common knowledge for decades.
In the 1980s, a friend of mine was working on his advanced degrees in mathematics. He originally proposed his thesis on Semantics as Mathematics in the 1970s. The universities thought he was nuts. Eventually, a Swiss (I believe) mathematician of some renown started working on a similar theory. The universities started trying to bring my friend back. His work ultimately contributed to technology such as texting and instant messaging - and that bit of AI you've been using (and arguing with_ for years), predictive text. We've all dealt with words being autocorrected to things like "DUCK" or "SNIT". Those are examples of biases in the program, or what the industry likes to call training data bias. Fun Fact: computer programs don't write themselves. They never will. An AI program might write the program faster and more efficiently, but like any program, it won't be able to do it without 'prompts' from the user, in this case a programmer.
I know this because it was explained to me 30+ years ago by another friend - with a PhD in computer science. He was working on some stuff, allegedly for the government, that sounded unethical to me at the time. (no, I won't go into that particular story) Mind you, I briefly studied programming in the early 1980s. I find computers fascinating, but I lack the attention span required to be any good at programming. That said, I understand the basics of it, and I understand the basics of AI.
It's little more than a very fancy database - a system trained on massive datasets that predicts patterns (like language or images) based on statistical probability. Again, it doesn't do this on its own. We, the users, have to prompt the program (whichever AI system) for the desired result. Often, I use AI for recipes! I'll type in a list of ingredients that I have available, and request ideas for a meal. Sometimes it works well, other times - not so much. From the few systems I've experimented with, there's usually a disclaimer that clearly states AI is capable of mistakes. In reality, this is usually due to the user's lack of clarity in giving prompts. Personally, I find some of the mistakes hilarious!
OK, is AI going to take your job? Probably not - but you might have to boost your own unique abilities. You have to go full-on John Henry! The reality is, AI has been around a long time. You use it every day. Musicians, artists, writers, teachers, etc. often complain that it is destroying their career potential! Again, no. AI can't CREATE. It can come up with an approximation, based on the prompts by the user, that might be all you need. For example, you want to write a paper on lesser-known battles in WW2? AI can probably do that. But it can't write a detailed interview with a survivor. AI is good for the quick stuff.
I've seen programmer friends use AI to help them write code. Hey, if it's a tool that increases speed and efficiency, I'm all for it.
But what about the creative world? The musicians, writers, artists, etc.? As a working musician since 1979, I can tell you this is just another hurdle. For my first handful of years in the biz, live music by live musicians was the way to go. By the mid 1980s, DJ's had infiltrated our business. Why pay an entire band when you can just hire 'Bob' to play human jukebox? Then karaoke started getting popular. 'Bob' didn't even need to bring crates of records. The audience became their own entertainment. I used to make my rent in one weekend. Nowadays - fuhgeddabowtit! Artists complain that no one hires them now. I hear this complaint mostly from artists who aren't being hired as artists in the first place. As a musician, I've been making posters and flyers since about 1981. Good ol' punk rock cut & paste! Unless we knew someone who could actually draw, paint, or was skillful with a camera, cut & paste is what we had. Of course, we could have hired an artist. Any decent artist would charge more than we were going to make - so it was a non-starter. In 46 years, the only times I've hired someone to do artwork of any kind was for promo photos, 1 t-shirt design (which we never used), and 2 album covers. Everything else, I've done myself. Am I especially creative? I don't think so - I just know what I want to see, and I find ways to make it happen. Real art is never about making money. It's nice when the money comes, but someone else is always going to make the lion's share - never the artist.
AI is here to stay, and it's not scary. It's just a tool. Like DJs and karaoke didn’t kill live bands; photos didn't kill painting; recording didn't kill live performance - AI won’t kill human creativity. It just changes the stage we’re playing on. It's prone to make mistakes. Some of the output is laughable. Some is amazingly good and accurate. I can almost always tell if something has been 'created' by someone using AI (as a tool). It tends to be boring and predictable. But occasionally, it does give me a good recipe that uses up that leftover chicken, potatoes, and whatever else I have in the fridge.
No comments:
Post a Comment