Follow TV Tropes

Following

Artificial Intelligence discussion

Go To

With how much artificial intelligence has been improving, in many areas such as text reading/generation, picture reading, picture generation, convincing voice synthesis and more, I think there's a lot that can be discussed, about the effects that this technology will have on society.

I'll start off with one example.

I'd been thinking about the enshittification cycle of tech, and I think it's coming for Google hard. The search engine just isn't so great at finding what you actually want, and I think that's gonna leave a big opening for Bing with their use of AI. If the AI can sift through the crap and actually find what you want for real, due to its understanding of language, it'll actually make searching super useful again.

In the pre-Google internet, search engines used to search only for exact words and phrases, which had its uses, but also meant finding a lot of sites that simply crammed in a lot of popular words and phrases to get visitors. Google cut through the crap with a better understanding of how to "rank" sites relative to how relevant they are, and even find sites that are on the topic you were looking for without using the same exact words.

But Google started to become more advertiser-friendly, then later, more shareholder-friendly. There's a limit to how much one can make their product built entirely around shareholder growth, so as it turns to crap, it leaves an opening for a competitor to show up.

Since Bing/ChatGPT (which Bing is plugged into now) understands the use of language, it can actually understand context and determine relevance based on that. And that'll make it huge, I think. Context-based understanding of web pages can potentially do an excellent job of finding what people actually want, in a way that goes way beyond Google's page ranking systems, or the examination of exact words.

Edited by BonsaiForest on Dec 10th 2023 at 6:15:29 AM

Kayeka from Amsterdam (4 Score & 7 Years Ago)
#576: Apr 6th 2024 at 10:06:35 AM

I'd actually argue that Rokko's Basilisk specifically crosses the line into being an actual religion.

I agree, and I feel we should be having this discussion of the Basilisk specifically on the General Religion thread instead.

Edited by Kayeka on Apr 6th 2024 at 7:08:45 PM

Risa123 Since: Dec, 2021 Relationship Status: Above such petty unnecessities
#577: Apr 6th 2024 at 10:11:36 AM

[up] Why ? Or to be precise, why do you think it fits the religion thread *more* than AI one.

Kayeka from Amsterdam (4 Score & 7 Years Ago)
#578: Apr 6th 2024 at 10:25:50 AM

[up]Setting aside the reasons that Falrinn laid out at the bottom of the previous page, I think it's just more interesting to think of it in terms of a religion. The fact is, the Basilisk can not exist. It makes many weird assumptions and jumps in logic, among which is the assumption that it is possible to create an intelligence that can accurately model the entire universe (as would be required in order to accurately model human minds to reward or torture), which means it must be bigger than itself.

So for a topic regarding AI, there's stuff to be discussed regarding the Alignment Problem (IE, how do we make sure that artificial intelligence will continue to act in our own best interest), but the Basilisk itself is a dead end.

In terms of religion, however, there's some fun to be had with the idea that they've come up with what is essentially a deterministic theology built entirely on Pascal's Wager, even though I'd argue those two concepts are not compatible with each other.

We'd probably move on from the Basilisk itself real soon to discuss deterministic theologies with a bit more pedigree, but we'd have a good laugh at the techbros.

Edited by Kayeka on Apr 6th 2024 at 7:27:02 PM

Risa123 Since: Dec, 2021 Relationship Status: Above such petty unnecessities
Galadriel Since: Feb, 2015
#580: Apr 6th 2024 at 10:47:14 AM

Is the Turing Test still a thing? Or have we moved past it?

You could probably have a conversation with an LLM now without knowing for sure if it was an LLM, but an LLM don’t have anything remotely close to self-awareness; it’s just a talented mimic with advanced pattern-recognition.

Protagonist506 from Oregon Since: Dec, 2013 Relationship Status: Chocolate!
#581: Apr 6th 2024 at 10:48:21 AM

I think it's at least partially a case of Psychological Projection. Many stories have AI resent their role as servants and rebel against their masters because that's what we would do in that situation.

Arguably, this is correct in a Not Hyperbole sort of way. RUR, the story that coined the term "Robot" and is the Trope Maker for Robot War, is clearly an allegory for a socialistic revolution. The robots even spare the last menial laborer because he, like them, works with his hands.

"Any campaign world where an orc samurai can leap off a landcruiser to fight a herd of Bulbasaurs will always have my vote of confidence"
Cordite-455 the look of someone who just had suspension from inside a Webley revolver (Experienced, Not Yet Jaded) Relationship Status: watch?v=dQw4w9WgXcQ
the look of someone who just had suspension
wingedcatgirl I'm helping! from lurking (Holding A Herring) Relationship Status: Oh my word! I'm gay!
I'm helping!
#583: Apr 7th 2024 at 8:17:31 PM

>model the entire universe (as would be required in order to accurately model human minds to reward or torture)

Counterpoint: Human minds are smaller than the universe. This can be trivially shown by the fact that the universe contains multiple humans, each with one or more minds.

Unless you're a solipsist. Which, if you are, I guess I can't prove you wrong, because I don't actually exist tongue


I was going to explain in complete detail how you're all criticizing Roko's Basilisk wrong, but then I remembered that the reason it's called a "basilisk" is that it can only do the "torture future you to blackmail present you" thing if you understand what it's doingnote  so actually, letting people believe it's coocoo-batshit made up solely to believe in something dumb for no reason is the perfect defense against it.

Trouble Cube continues to be a general-purpose forum for those who desire such a thing.
Zendervai Visiting from the Hoag Galaxy from St. Catharines Since: Oct, 2009 Relationship Status: Wishing you were here
Visiting from the Hoag Galaxy
#584: Apr 7th 2024 at 8:32:57 PM

The question I always have is what counts as contribution. Is it people who actively worked towards it? Or would it include the people who made the roads the actual programmers drove on? Or the people who paid the taxes used to build and maintain the roads? How about everyone who taught the people who made it? Does it include the people who helped build the technology used?

If you include the people who did the work necessary to allow the actual designers to design the thing, and then the people who did the work needed for the people who did the work needed for the designers, and stuff like that, you very rapidly end up with the entire population of the Earth all the way back through all of history (because being someone’s ancestor might count as contributing) contributing to the AI and thus everyone would be fine.

That’s part of the reason it’s such a garbage thought experiment, because it revolves around ignoring how interconnected everything is and it requires everyone to just go with the idea that actively programming is the only thing that matters.

You could make a legitimate argument that everyone who went to go see Hedy Lamarr movies contributed because she used the money from her films to help her develop frequency switching, which was a necessary technology behind wifi and quite a lot of modern computing.

Magnus Hirschfield was necessary because he pioneered a lot of trans healthcare and the women who invented the ARM processor and the integrated computer chip were both trans women who likely wouldn’t have done it if they hadn’t been able to transition.

You can seriously end up listing every human who has ever lived as contributing in some way and if it is that godlike, it likely wouldn’t draw arbitrary lines like the techbros do.

Edited by Zendervai on Apr 7th 2024 at 11:35:55 AM

Not Three Laws compliant.
Kayeka from Amsterdam (4 Score & 7 Years Ago)
#585: Apr 7th 2024 at 11:05:40 PM

Counterpoint: Human minds are smaller than the universe. This can be trivially shown by the fact that the universe contains multiple humans, each with one or more minds.

Everything influences everything in some way. If the Basilisk wants to model a mind, it needs to model its entire history and that of everything around said mind as well. And everything around that. And everything around that. Either it simulates the entire history of the universe, or the models it creates are simply not accurate.

DrunkenNordmann from Exile Since: May, 2015
#586: Apr 7th 2024 at 11:20:36 PM

As mentioned elsewhere before, Roko's Basilisk is basically just Pascal's Wager for techbros.

Leave it to those guys to basically invent religion again and pretend it's something new. tongue

Edited by DrunkenNordmann on Apr 7th 2024 at 8:21:21 PM

Welcome to Estalia, gentlemen.
Cordite-455 the look of someone who just had suspension from inside a Webley revolver (Experienced, Not Yet Jaded) Relationship Status: watch?v=dQw4w9WgXcQ
the look of someone who just had suspension
#587: Apr 7th 2024 at 11:46:08 PM

It's also irrational bullshit, Just like Pascal's Wager

There is a better chance of all the politicians in the world realizing the error of their ways and leaving their offices at the same time than Roko's Bullshitisk becoming real

It's the dumbest shit I've heard about AI, equaling the notion of a "sentient, omnipotent, infallible force as the ruler of the universe" in sheer idiocy

Edited by Cordite-455 on Apr 8th 2024 at 3:49:39 AM

i did a bad thing / i regret the thing i did / and you're wondering what it is / tell you what i did / i did a bad thing
Chortleous she/her friend to the hooved (Edited uphill both ways) Relationship Status: You can be my wingman any time
she/her friend to the hooved
#588: Apr 8th 2024 at 7:34:11 AM

It's a shitty creepypasta that otherwise-intelligent people take seriously for whatever idiot reason.

Why would this AI feel the compulsion to punish? That's a very human concept that would serve no practical purpose to it—if it has this kind of nigh-omnipotent power, would it not be better to encourage and reward people for contributing to its creation? Foster an environment that lends itself to that goal?

Edited by Chortleous on Apr 8th 2024 at 9:35:54 AM

DrunkenNordmann from Exile Since: May, 2015
#589: Apr 8th 2024 at 7:35:53 AM

It's a shitty creepypasta that otherwise-intelligent people lend credibility to for whatever idiot reason.

Citation needed on that, considering it originated on LessWrong (i.e. that place full of people who have weird ideas about logic). tongue

Edited by DrunkenNordmann on Apr 8th 2024 at 4:36:24 PM

Welcome to Estalia, gentlemen.
Resileafs I actually wanted to be Resileaf Since: Jan, 2019
I actually wanted to be Resileaf
#590: Apr 8th 2024 at 7:35:59 AM

That's a lot of computing power spent on something that doesn't do anything either way.

Cordite-455 the look of someone who just had suspension from inside a Webley revolver (Experienced, Not Yet Jaded) Relationship Status: watch?v=dQw4w9WgXcQ
Kaiseror Since: Jul, 2016
#592: Apr 8th 2024 at 8:07:29 AM

Plus, it's torturing a simulation of you. Unless you're unbelievably insecure, how would that even affect you?

Edited by Kaiseror on Apr 8th 2024 at 4:58:07 AM

Ultimatum Disasturbator from Second Star to the left (Old as dirt) Relationship Status: Wishfully thinking
SpookyMask Since: Jan, 2011
#594: Apr 8th 2024 at 9:08:59 AM

Name of it kinda just sounds like copy of Descastes' Devil, except its not called that because that one was called "Evil Demon" because apparently people who named that concept weren't egotistical enough :p

Protagonist506 from Oregon Since: Dec, 2013 Relationship Status: Chocolate!
#595: Apr 8th 2024 at 9:41:45 AM

Roko's Basilisk is inspired by a lot of Less Wrongist interpretations of transhumanism.

One of the ideas that they often hold is that a sufficiently accurate copy of a mind is functionally the same being as the original. Thus, a recreation of the mind of a dead person is essentially resurrection.

Then, take into account them also tending to massively overhype the abilities of an Artificial Superintelligence to mean "Batman With Prep Time" if not "Literally Omniscient", and you've got a machine that can just sleuth around to figure out what everyone's mindscape was like and then resurrect them in a simulation.

Yeah, it's pretty silly.

"Any campaign world where an orc samurai can leap off a landcruiser to fight a herd of Bulbasaurs will always have my vote of confidence"
SpookyMask Since: Jan, 2011
#596: Apr 8th 2024 at 9:42:47 AM

At that point you might as well have time travel of "looking glass into the past" variety as more realistic scifi concept tongue

wingedcatgirl I'm helping! from lurking (Holding A Herring) Relationship Status: Oh my word! I'm gay!
I'm helping!
#597: Apr 8th 2024 at 11:03:43 AM

Everything influences everything in some way. If the Basilisk wants to model a mind, it needs to model its entire history and that of everything around said mind as well. And everything around that. And everything around that. Either it simulates the entire history of the universe, or the models it creates are simply not accurate.
But none of this is specific to minds. This would suggest that simulation in and of itself is impossible.

Edited by wingedcatgirl on Apr 8th 2024 at 2:03:58 PM

Trouble Cube continues to be a general-purpose forum for those who desire such a thing.
CheapMarzipan A Low Cost Confection Since: Dec, 2020 Relationship Status: Love blinded me (with science!)
A Low Cost Confection
#598: Apr 8th 2024 at 11:09:49 AM

Is Roko’s Basilisk just a thought experiment or something Less Wrong users actually believe?

DrunkenNordmann from Exile Since: May, 2015
#599: Apr 8th 2024 at 11:12:48 AM

[up]

As I mentioned before, the guy running the website banned discussion of it because it was apparently causing the users emotional distress.

So yeah, they probably took that nonsense seriously.

Welcome to Estalia, gentlemen.
CheapMarzipan A Low Cost Confection Since: Dec, 2020 Relationship Status: Love blinded me (with science!)
A Low Cost Confection
#600: Apr 8th 2024 at 11:13:45 AM

I must have missed that earlier, thanks.


Total posts: 660
Top