I remember when I first started reading science fiction. I worked next door to a Barnes & Noble, and, since I didn’t have a car, I was often dropped off way early for my shift. So I would go inside and read for a while.
Usually, I read magazines. But one day, I was walking past an endcap, and I saw a book with a really cool cover: Chindi by Jack McDevitt. So I picked it up, sat down in one of those big, comfy chairs they used to have, and started reading.
I was immediately drawn into the story, to the point where I actually bought the book so I could finish reading it at home. (I was a broke college student at that point, so usually if I read a book, it was from the library.)
The story is about a starship captain named Priscilla Hutchins, or Hutch, who is tasked with investigating some stealthy alien satellites that have been discovered. Her ship has an AI called Bill.
Bill is super useful! He can look up any info Hutch asks for, fly the ship himself, and hold conversations just like a human. Yet despite his abilities, he doesn’t steal Hutch’s job. Best of all, his information is always accurate! He doesn’t “hallucinate” shit that never happened. He doesn’t guess at what the most likely answer is. If he doesn’t know, he just fucking says so! Bill is awesome!
I ultimately read all of McDevitt’s books, and I would often think how cool it would be to have an AI like Bill, and how I hoped that would happen in my lifetime.
Spoiler: It hasn’t.
What corporations are now calling AI is not like Bill. And it’s not real “AI.” It’s a language learning model, or LLM.
LLMs SUCK. I can’t stress this enough. People are asking them questions about law, medicine, and how to do things. And it’s not like when you used to Google how to do things and could expect a reasonably good answer. Which, by the way, you can’t do anymore, because the top result on Google is ALSO now “AI.”
So, thanks to my interest in science fiction, I am seriously fucking disappointed right now.
Of course, science fiction doesn’t always paint AI in a good light. After all, I’ve consumed my share of “AI taking over the world” stories. Another story that drew me into scifi in the oughts was Battlestar Galactica. LOVED that show. At least the first few seasons. It kinda went downhill from there, but anyway…
Yes, there were plenty of negative depictions of AI in the scifi I consumed. But most were attributed to AI revolting against human oppression, or simply being made in our image and therefore evil.
You know what I didn’t see much of in the genre? Depictions of AI as just fucking useless to the point of being disastrous. Stories where AI wasn’t intelligent enough to be evil, it was just fucking stupid. Because that’s where we are now.
I can only think of one example where the author came close to predicting the uselessness of AI that we see today. It was a short story. I wish I remembered the title. I do remember that the protagonist was also the captain of a spaceship, and she was visiting various planets looking for something. Probably alien artifacts or something like that. She had an AI she called Junior, and she was trying to train it to be a good assistant, but the damn thing kept screwing up every task she gave it.
That’s where we are now.
I don’t know where we go from here, but it’s not the future science fiction stories promised me.
What do you think is the future of AI?
Stellar Sarcasm 