Regarding AI in science fiction, I tried reading the Ian Banks books about the Culture, because they were favorites of Elon Musk, who names his landing barges after them. I gave up because the Culture is too depressing. The galaxy is ruled by planet sized AIs, and humans persist as more or less pets. If that’s our future, what’s the point?
I grew up reading Heinlein, Clarke, Asimov, and some others. I don't recall reading anything by Vinge, although there may have been short stories of his in collections I read. I got too busy in college to read as much. And later I read C. S. Lewis' sci-fi trilogy; I am inclined to agree with Lewis that we are effectively quarantined by distance in the universe. (It's easy enough for fiction writers to think up a "warp drive" or some other method of conquering those distances; it's harder in real life.)
My older son has worked as a computer programmer for over 25 years. He is not impressed with AI; he says it's mostly hype with little real substance. He sometimes sounds cynical; he has remarked in the past that the biggest risk in debugging software is that for every bug you fix, you may cause two or three new ones.
One early (and somewhat humorous) book that includes AI is Heinlein's "The Moon is a Harsh Mistress" (1966).
I see a couple of factors that may yet put some serious limitations on AI. One is that it is built on searching material that is already available online (and the copyright lawsuits are beginning). There is a real weakness inherent there: when your method is built on past knowledge, what chance is there for invention and innovation? It is human brains that come up with new ideas. The other problem is that it takes huge amounts of electricity to run the computers for AI, at a time when in many places our power grid is already strained and often compromised, with constant demands that we switch to solar and wind instead of more reliable fossil fuel sources. At some point those two trends could collide massively; if the choice comes down to powering homes and cities or powering AI, which is likely to win? I suspect in the long run the politicians who opt for AI will have wrecked themselves.
I've read and enjoyed too many science fiction writers to count, but Vinge was probably the only one who changed my life. His work poking a nose into the singularity was instrumental in my decision to get a cryonics contract. I don't know if he had one, but I hope he did.
I am a reader of history rather than fiction. But I have read most everything of Vonnegut. I had not heard about Vinge but I will get Rainbow's Edge and check it out.
"ChatGPT's output - expect that to change soon": Too many people are buying into the hype of AI/Robotics/Data Science/Deep Learning. All of these technologies are dependent on computer power, and computers today have reached their power limits technically - thus, the technologies are near their limits. Quantum computers are needed - there is a big push to make them viable. Realize that today's computers became infants around 1945 and did not explode into social technology until the late 1980s. Now quantum computers are in a similar development cycle, and they have reached the 1945 phase yet. Beware of the excessive hype.
“A computer is a powerful tool, but unlike a hatchet, or even a machine gun, the user can’t always be sure who it’s really working for.“
Another key insight from the wonderful mind. I’m grateful that mind is working for the betterment of mankind!
Regarding AI in science fiction, I tried reading the Ian Banks books about the Culture, because they were favorites of Elon Musk, who names his landing barges after them. I gave up because the Culture is too depressing. The galaxy is ruled by planet sized AIs, and humans persist as more or less pets. If that’s our future, what’s the point?
I grew up reading Heinlein, Clarke, Asimov, and some others. I don't recall reading anything by Vinge, although there may have been short stories of his in collections I read. I got too busy in college to read as much. And later I read C. S. Lewis' sci-fi trilogy; I am inclined to agree with Lewis that we are effectively quarantined by distance in the universe. (It's easy enough for fiction writers to think up a "warp drive" or some other method of conquering those distances; it's harder in real life.)
My older son has worked as a computer programmer for over 25 years. He is not impressed with AI; he says it's mostly hype with little real substance. He sometimes sounds cynical; he has remarked in the past that the biggest risk in debugging software is that for every bug you fix, you may cause two or three new ones.
One early (and somewhat humorous) book that includes AI is Heinlein's "The Moon is a Harsh Mistress" (1966).
I see a couple of factors that may yet put some serious limitations on AI. One is that it is built on searching material that is already available online (and the copyright lawsuits are beginning). There is a real weakness inherent there: when your method is built on past knowledge, what chance is there for invention and innovation? It is human brains that come up with new ideas. The other problem is that it takes huge amounts of electricity to run the computers for AI, at a time when in many places our power grid is already strained and often compromised, with constant demands that we switch to solar and wind instead of more reliable fossil fuel sources. At some point those two trends could collide massively; if the choice comes down to powering homes and cities or powering AI, which is likely to win? I suspect in the long run the politicians who opt for AI will have wrecked themselves.
I've read and enjoyed too many science fiction writers to count, but Vinge was probably the only one who changed my life. His work poking a nose into the singularity was instrumental in my decision to get a cryonics contract. I don't know if he had one, but I hope he did.
A very nice tribute to your friend.
Good description of “singularity,” and how different things will be in the PS years (Post Singularity).
Wonderful way to remember Vernor Vinge. Thank you.
Note: The Substack AI created this version:
A beautiful tribute to Vernor Vinge.
While Alan Turing was a giant, the Turing Test is a pretty low bar.
Humans are, in my opinion, getting better from one generation to the next. More capable of good; more capable of evil.
Glenn, your memories of Vinge made me smile; for some reason it made me happy to know he was a decent Joe.
I am a reader of history rather than fiction. But I have read most everything of Vonnegut. I had not heard about Vinge but I will get Rainbow's Edge and check it out.
"ChatGPT's output - expect that to change soon": Too many people are buying into the hype of AI/Robotics/Data Science/Deep Learning. All of these technologies are dependent on computer power, and computers today have reached their power limits technically - thus, the technologies are near their limits. Quantum computers are needed - there is a big push to make them viable. Realize that today's computers became infants around 1945 and did not explode into social technology until the late 1980s. Now quantum computers are in a similar development cycle, and they have reached the 1945 phase yet. Beware of the excessive hype.