I have a hard time suspending disbelief with sci-fi these days. It's really hard to imagine any believable future along current trend lines; either you contrive a reason why AI didn't become a thing (1), or you have to explain why humans are doing anything - piloting a starship, science, exploration, etc. (2)

(1) Butlerian Jihad. Sanderson's Skyward at least had a reason.

(2) Banks's Culture had a possible future. I'm sure there are others, but not many.

I think about this a lot, and while I have tons of near future sci-fi ideas (which probably would have seemed very far future, 10-20, years ago), I just can't conceive of anything much farther out. I guess that's why they call it the "singularity".

AI from the past seems more like it’s based on the computers of the time - the AIs seem to be very logical, never make mistakes and act like they’re an expertly programmed bunch of switch statements and ifs. Whereas modern AI is based on human text and so the models have human biases and do things like swear in their chain-of-thought and sometimes make errors for no reason at all.

Not mentioning Peter Watts here is criminal, because he came unnervingly close to nailing it. The voice of Rorschach in Blindsight is very similar to early LLMs. And, well, hopefully the dead internet of Starfish doesn't turn out to be a prophecy but I'm getting nervous.

I’ve been thinking about this for awhile now, a complete miss: Sci-fi author and computer scientist Vernor Vinge popularized the term programmer-archaeologist in his 1999 science fiction novel A Deepness in the Sky

The premise is that AI never came around and we were left scavenging code since that was more efficient than writing new code. But with LLMs, it’s much easier to write new than to scavenge.

Nothing before 2017? What about:

  - Sally, Asimov 1953 (spot on with Waymo)
  - I, Robot etc, Asimov
  - HAL 2001 A Space Odyssey
  - any Iain M. Banks Culture novel

Many sci-fi writers really dislike LLM for various reasons, whether it is the generative works they produce, or for allegations of copyright infringement due to LLM training on their works. Some have taken such a negative attitude towards LLM that they actively refuse to use Claude or GPT even for coding projects. John Scalzi and Catherine Asaro come to mind boycotting any LLM use…

I'm glad to see Aurora (by KSR) up there. It's a great sci-fi story overall, but for this topic the "waking up" of the AI and the entire evolution as we get to know it, and hear its voice is absolutely great. Highly recommend it, if you haven't read it.

The sci-fi story that I constantly find myself thinking about when I read about real-world AI developments is Ra, by qntm/Sam Hughes: https://qntm.org/ra

There's also "Destination: Void" by Frank Herbert.

Iain M Banks' Culture series seem to have got a lot right.

Neal Stephenson's Diamond Age (a young lady's illustrated primer) ≈ LLMs + Intelligent UI (OpenAI)

vernor vinge bubble in time series gives a lot of food for thought without overly patronizing or moralizing

This doesn't seem accurate at all. E.g.

> A high-scoring story shows readers moved by machine literature, editors publishing it, an audience enjoying its music, or critics taking the work seriously.

How is this a criterion for LLMs being sci-fi AI? If anything, opposite is true. It seems that the author is comparing his own sci-fi image of LLMs with sci-fi. Or Claude's vision of itself, I guess, since this is all vibe-written except the initial few lines.

This goes so hard with numbers without even a passing mention of how the author actually read the books. AI wrote this.

This question assumes "this is it, this is AI", which I find laughable. A necessary stopgap accelerator at most, I think people will look back at LLMs like toys in the future, and our societal bewonderment as a result of mirage.

I think the score for Colossus: The Forbin Project is too low. There is nothing in the film that contradicts the notion of Colossus being a very, very large LLM. Although I think the film is even better than the book by D. F. Jones, only the latter mentions how, despite being created specifically for US national defense, Colossus is also fed unrelated data including Shakespeare's sonnets, because its creators do not know if it could be important.

>The two machines begin with arithmetic and soon communicate through mathematics their engineers can’t follow. That’s the most relevant prediction here: machine intellectual activity racing ahead of the people supervising it.

This misses the specificity of the scene's prediction. Colossus and Guardian are basically speaking in the style of the super-dense, almost incomprehensible-to-human chain of thought that LLMs indulge in nowadays. One might also point to the just-released news of LLMs easily seeing patterns, incomprehensible to humans, in model weights.

That jump in computer intelligence, so soon, is completely surprising. And I bet most actual sci fi fans would agree.

I used to dismiss anything near-future involving any robot or AI you could just talk to. Now it exists. Unfathomable.

What AI?

It's interesting how the Hard SF authors (at least, the ones I read) had poorer predictions than the more fantasy-like ones. Asimov, who was considered a giant of the field, thought that AI (robots) would have difficulty with ambiguity, creativity, and lying. This trend extends all the way to Star Trek's Data. Looking back on it, this view was probably based on Asimov consciously or unconsciously basing his robots on the more autistic of the scientists that he knew.

Whereas fantasy authors write stuff like "a sufficiently large pile of books becomes an intelligence" (Pratchett) and that's pretty much what's happened...

It's gonna be cool to reevaluate that in few years. Different stories might move to the top as real world AIs capabilities grow and new behaviors manifest.

Everybody needs to go read Asimov's Someday. Now's a good time, go do it.

I think this quote is cherry-picked:

> I agree with Ursula Le Guin that “Science fiction is not predictive; it is descriptive.” To me, the interesting parts of science fiction isn’t how accurately it extrapolates into the future, but rather, given a future circumstance, how might be?

You can read her views at https://www.penguinrandomhouse.ca/books/342990/the-left-hand... .

"""Prediction is the business of prophets, clairvoyants, and futurologists. It is not the business of novelists. A novelist’s business is lying.

The weather bureau will tell you what next Tuesday will be like, and the Rand Corporation will tell you what the twenty-first century will be like. I don’t recommend that you turn to the writers of fiction for such information. It’s none of their business. All they’re trying to do is tell you what they’re like, and what you’re like—what’s going on—what the weather is now, today, this moment, the rain, the sunlight, look!"""

Le Guin's view, as I understand it, is that science fiction is about describing the present, even when it takes place in the future.

I spot-checked "The Moon Is a Harsh Mistress". It ranks a 2 for scaling, but according to the guidelines it should rank 0, because "A giant computer that merely stores more records doesn’t count for much. Neither does a magical threshold where the machine suddenly wakes up." and Mike's description is literally that he woke up when he reached that threshold.

It's on the first page. https://archive.org/details/bwb_S0-DFY-783/page/8/mode/2up?q...

The growth in Mike's abilities is not from putting new hardware online, but because he matures like a person does, for example, https://archive.org/details/bwb_S0-DFY-783/page/64/mode/2up?... , which also describe the even as "woke".

I don't see how one can read the book and miss that.

I also don't like the omission of Wyoh's contribution to Mike's training.

Sci-fi writers were generally wrong or over-simplistic about many predictions, mostly just extrapolating whatever the current zeitgeist of their time was, and stretching it out to 50 or so years into the future

I mean look how most science fiction of the 1980s depicted the world of the future as dominated by Japan... Look at Robocop 3 :)

Computers were then just appearing or increasing in public discussion, along with fears that they'll make humans less human somehow: Like in this Computer Chronicles video about word processing from 1983 at 4:08: "Some writers are concerned that computer assistance may promote dry bland writing."

https://www.youtube.com/watch?v=Jt0OoXluC8g

AI predictions were generally hostile to humanity: Terminator appeared in 1984, and the first Alien movie in 1979 had an antagonist android. 2001 A Space Odyssey with a disagreeable AI came out in 1968. All way before the real world had anything remotely close to today's AI: It was a reaction to the appearance of computers themselves.

On the other hand, Star Trek had optimistic AI like Data.

Humankind's own sentient creations leading to the ruin of humans is probably a trope in fiction since thousands of years, with golems or summoned demons etc.