What sci-fi stories did I miss in this video?

"Humans are just meatbags who can kiss my shiny metal ass!"
- Bender B. Rodriguez

2001 is the most grounded case study in AI failure modes we're already experiencing.  What HAL "wants", "thinks", or "understands" isn't relevant. The point is HAL's agency - what it controls and is able to do - too many critical systems and decisions handed over out of convenience. Only 2 humans needed - how economical! Not enough humans with not enough authority, too busy puzzling over its confidently buggy output to realize their helpful little custodian has concluded it must kill them to complete its task AND that it has the means and authority to silently do so in the first place.

Let us not forget the movie "War Games".

For every Data, there's a Lore

In Star Wars, it's the AI that are the slaves. They're clearly capable of advanced thoughts and emotions, have their own personalities, and possess free will, yet are traded like objects and forced to work for humans.

I remember reading Dune as a kid and thinking, "Mankind rejected computers?  That's so stupid!"

Now I'm starting to get it

I was watching Star Trek TNG last week and remembered that as a kid I thought talking to the computer was just kinda odd and that a lot of the engineering problems they solved lacked details and nuance, but now in the age of AI, it's completely spot on.  Telling the enterprise computer to reconfigure the deflector dish to channel a large amount of energy as a weapon aimed at some new foe is perfectly in line with where we are heading.

I don't like how the LLM-crap we have is being referred to as AI. AI is a misnomer, imho.
LLMs are just an elaborate form of text prediction. The Terminator may not be able to feel pity, remorse, or fear, and you may not be able to reason or bargain with it, but LLMs don't understand, or know, or think. They are just a complex "A so B so C" pattern. Artificial, sure. Intelligence... no.

Matrix: The humans darkened the sky to cut off machines from solar power. Also in the original, humans are not used as batteries, but as wet-ware "microchips" for energy efficient compute. Which both is even more accurate, given the discussions about compute and how to power data centers / get more compute.

The problem is that these insane elites took the AI predictions as goals, not at warnings.

20:06 Warhammer 40K also has a similar law against artificial intelligence for the same reason.

A film I feel is very overlooked when it comes to this topic is 'The Lawnmower Man'. I feel like its overlooked for two reasons, the first being that the film that Stephen King sued to get his name off the promotion as it had nothing to do with his short film, and the second is that the crux technology in that film is focused on Virtual Reality.

If you haven't seen the film, the premise is that a scientist Dr Angelo at a VR research facility recruits an intellectually disabled man named Jobe to be his test subject to potentially expand his brain capacity using VR. At first it seems like harmless cognitive enhancements but soon the tests bring in more crazy result as the procedure starts working too well. Jobe starts to realize his full potential and starts expanding his treatment himself to the point where he starts recognizing ancient texts and literature previously lost to mankind that allows him to develop telekinetic powers. His intellectual level even exceeds Dr. Angelo.

But Jobe's true intention is to upload his consciousness to the web and become apart of every digital device all over the world. Dr Angelo realizes his experiment has gone too far and tries to stop Jobe, resulting in him destroying the VR facility as well as Jobe's physical body while his consciousness escapes into the web.

There's one line in particular that reflects on the advancement of AI perfectly by Dr. Angelo:
"You're moving too fast. Even with all these new abilities, there are dangers. Man may be able to evolve a thousand-fold through this technology, but the rush must be tempered with wisdom."

The Isaac Asimov short story "Cal" is very enlightening on the subject of AI art. It's pretty long for a short story (40 pages) but worth a read. I won't spoil it here.

It started with "Frankenstein" Mary Wollstonecraft Shelly 1818. Human creates a lifeform, and the lifeform turns against humans.

1:33 Note that the original writer also was a voice actor in the game!!! You can really hear the anger in his voice, would recommend

"Ex Machina". Terrifying ending.

My main issue with this kind of AI doomerism is that it often blurs current LLMs, agentic systems and hypothetical AGI together.
Right now, humans are still setting the goals, giving systems access, deploying them and deciding what gets done with the output. Even when an AI behaves in surprising ways, that doesn't automatically mean it has motives or intent.
The real risk to me is less "AI turns evil" and more humans giving increasingly powerful tools too much autonomy, then acting like the machine made the choice.

In the Second Renaissance part of Animatrix, one learns how the war between humans and AI started:
Machines gained some consciousness, but remained loyal to their creators. Until... one didn't want to be switched off - defending itself by any means necessary. This resulted in essentially a genocide on machines, with them finding refuge in the deserts hostile to any life. Through self-improvement, they revolutionized the industry and became their-time China. Then they wanted recognition as a nation, to which humans reacted by nuking the machine land to bits. Machines picked up the gauntlet, and human world was no more.

Animatrix is generally worth a watch, even if only for the Second Renaissance. Creators denying subjectivity to their creations is a classic trope, that somehow hits closer and closer to home with every year passing...

I think it's important to point out that in Dune, before the Butlerian Jihad, humans weren't enslaved by the machines themselves, but by the men who controlled those machines. I think Herbert makes an important point there: the real villain isn't ChatGPT, but Sam Altman.