Vol. 89, February 27th, 2024 Published a day early online
The Burning Restaurant
The Boy Who Cried Wolf did it wrong: he didn’t have a TV show. While it’s always been fashionable to yell about the end times, the modern era concerns me. I don’t see how 2024 can’t be a rollercoaster (at best). The only agreement I’ve heard is that something smells like it’s burning.
So what to do about it?
A bootleg copy of Havel’s The Power of the Powerless showed up, howling at my door. Written under extreme persecution in Communist Czechoslovakia, it proposes (among other things) an answer to a pathological society is to refuse to participate in the lie. One truthful person shatters the illusion that the “game” is real.
I transpose this to our situation today: We sit in a burning restaurant, arguing bitterly about the menu. “This nothingburger is our ruin! Go vegan!” some say, while others blame the cooks for the flames. A brave man stands up, notices an exit door, and walks out into the fresh air, in search of what he knows to be true.
Havel illustrates the “pre-political” as the necessary starting point: the search for the answers to good living now, and that starts with refusal to play a game of lies.
It’s an essential read, and our book of the week.
Quote of the Week
“Individuals can be alienated from themselves only because there is something in them to alienate.”
–Vaclav Havel
Happy Birthday, Symphony No. 8
Premiering today in 1814, Beethoven was asked why it was less popular than his 7th.
“Because in all subtler respects, it’s better.”
Carol’s Appalachian Word of the Week
Carol Stuart sends along this “chestnut.”
Chester: Chest of drawers. “I bought a new chester for her bedroom.”
Signs of Spring
The trees are starting spring in Rustburg. A silver maple blooms on a sunny morning on the mountain. All around, the hills take on a fuzzy look as trees bud out.
Write to Us!
The Nighthawk is a new old-fashioned way to connect, published weekly. You’re invited to write back, or just enjoy reading. Let’s have some fun! It’s a social paper! Send stories, etc to: PO Box 783, Rustburg, VA 24588 or JoshUrban@protonmail.com
Letters from Josh
(A weekly update from Josh Urban’s adventures on the farm and in the city. #175)
Late Night Radio–With Josh Urban
The Robot Dinosaurs and Their Darn Suspenders
Howdy, folks, and welcome back to the show! Whatcha think of that Artificial Intelligence (AI) that's been sweeping the news? I made an account this morning, and typed “Draw hipster dinosaur in a diner.” The computer (or my parents) must have a sense of humor, because it spit out a picture of a T-Rex that looks like me. Maybe it's the suspenders. I jumped.
I've noticed something curious about the AI topic. Nobody seems to understand it fully. Maybe because we can't, and partly because we won't. Without further study, people start spouting a wide variety of opinions: It's the answer. It's the end. It's ruining art. It's biased. It's smart. It's stupid.
The jump goes from what is this to what does it mean for the future so quickly on any new topic, that predictions are often hilariously wrong. Have you seen (on the Internet) Nobel prize-winning economist Paul Krugman saying “the Internet’s impact on the economy has been no greater than the fax machine’s….ten years from now, the phrase 'information economy' will sound silly.”
So much for experts. So...what's the deal with AI?
Oh boy. For starters, the question is teaching me to think. When I'm faced with a problem, especially an urgent one (unexpected company in ten minutes), I do the “rational” thing: throw up my hands, and run in circles, sort of like a roach. But louder. It's not productive.
In thinking about AI, I'm trying to: 1. Understand the topic before trying to understand the future. 2. Be Skeptical. It's nice to read one thing and be done with it, but–we've all seen how unreliable single-source “experts” can be. (This absolutely includes the media.) I like a few sources, and I like to ask questions. 3.Try it for myself. ChatGPT is fascinating. 4. Talk with others about it.
Here's what I've found so far. AI is a lot of things: the auto-complete on your text messages, the robo-teller on the phone, a chess robot. The AI that's riveted the attention of the world comes in a lot of flavors, but ChatGPT is the rockstar. In Josh lingo, the computer “ate” billions of pages of writing, and learned the patterns of human communication. When I ask it to write a story, it'll build one from scratch, based on patterns that have a probability of working, not stored text. (It's not going to a special little shelf and picking out “poems about dinosaurs with suspenders that look curiously like the author.”)
How does it do that? There's a variety of ways, but a term you'll hear is neural network. With a system similar to our own brains (although much simpler), and some human input during training, it “learns” the correct way to approach the information through random trial and error. When a correct outcome happens, the pathway through the network is reinforced, while a bad outcome weakens the path just tried. It eventually learns enough to function on its own.
A toddler's brain does the same thing. “Oo, look at that red shiny thing. Should I touch it?”
If he does, and it's the red-hot stove burner, his brain will learn quickly that's an error, avoiding things that “look hot” in the future.
Still, the topic is befuddling, and it's easy to jump to panic that the robots–or hipster dinosaurs– are on the march.
So, I'm hosting a community talk at the Timbrook Library in Lynchburg on March 7th at 6 pm. It's free, so come on by and let's get into a fun discussion. I've got more details, some examples that make sense..and dinosaur pictures. We're going to have a great evening. Hope to see you there for some good human conversation.
Catch you on the flip side,
Josh
Send correspondence and smarties to P.O. Box 783, Rustburg, VA 24588 or on X @RealJoshUrban