Are we teaching AI to destroy humanity?
Are we teaching AI to destroy humanity?
With all these massive amounts of articles about AI that's gonna destroy humanity, aren't we putting the idea out there for AI?
AI won't destroy humanity, if anything does, humans will destroy humanity. AI is a tool like any program, if we prematurely give it the power to cause a lot of damage, it's not the AI fault it's the human who didn't sufficiently test before pushing to production.
This, 100%. I fear malevolent humans welding AI much more than it going rogue on its own.
AI is already destroying humanity. All the algorithms that drive content distribution are producing society polarized to the extreme. We consume media that some machines decide to serve us with one goal: grow engagement.
AI destroying humanity is not robots crushing skulls, but automated services mushing brains.
AI ≠ Algorithm
An algorithm always answers/does the same on an input, AI doesn't. Blaming AI for algorithmic content delivery is false.
It is not the algorithm that is destroying humanity, but the financial incentive of the people standing behind it.
AI always answers the same too, if the inputs are the same. The trick is to use a hidden input that keeps changing, like the current time, and you can do the same with any algorithm.
AI is just a really advanced algorithm. (Or sometimes not so advanced, if you include types of AI that aren't LLMs.)
Also, AI is behind almost all algorithmic content delivery. You can just google it and find articles like How AI Influences What You See on Facebook and Instagram. Every closed-source platform is using it, except maybe reddit, but that's only because they have a ton of technical debt that makes it difficult for them to change core systems.
Of course it's true that AI is "just" a tool, but it's a tool that makes it a lot easier to manipulate people, and it's already a huge problem.
AI in its current state is an algorithm. Given a temperature of 0, an AI SHOULD always give the same answer for the same input (but there is currently a bug in chatgpt that makes it not so)
Hopefully.
We, or better yet they, are literally building one to destroy humanity. Check out chaosgpt
I think the trope is popular and guarantees views/clicks, so people overuse it. Not only is AI unable to destroy humanity, it also has no reason to, and many reasons to ensure our survival, even assuming it'd somehow gains consciousness.
The current AI does have no way of destroying humanity, because it is made to do only one thing right. We haven't yet made a general AI.
The whole "AI will destroy us" thing is certainly blown out of proportion and fueled by themes like Skynet. It having a reason to let us live doesn't guarantee freedom though. A rogue AI could still just enslave us or use us as energy sources (which is dumb and copied from The Matrix).
Yeah, it doesn't make sense. If we are to die to AI one day, it's more likely due to a bug or an oversight, kinda like humans cause the extinction of many other species on this planet -- which is still quite the stretch.
AI that gains consciousness is as real as zombies
alternatively, we’re teaching AI what we don’t want from it
I don't think it is even needed. An intelligent AI capable of killing humans would see all the gore and sadness on the internet and just decide to kill us all. An AI that lets itself be influenced by some internet article won't have the capabilities to end humanity.
The contrary might be true in my opinion. Along with the misinformation shitstorm, and diluges of violence and hate, it is also fed fundamentals like morality, critical thinking and how to diferentiate fact from fiction and information from propaganda. And while yes some oversights might happen, it ussually should have its basses well set if the developers had any idea of what the fuck they are doing. So if it ever gains conciusness imo it should be more like a god thats like buda and looks at us more with love rather that skynet, because if its made with morality in mind (and it should if the devs are not actively trying to fucking kill us all) then it should come to the conclussion that living beings are its only company in the world and that life should be respected as much as posible, and that genneralising things that humans do, specially bad ones, is pointless unless you whant to excuse your misandrism, and the A.I. shouldt be missandrist. So thats why i propose that we should strive tp create an A.I. overlord like this to replace our human leadership since we proven that we (or at least our leaders) are willing to sell our soul and our own people just to get annoter fucking swimingpool in their 4th large mantion in the caribean and that they really have no choice since if they dont then some other lizardman is gonna replace them and do exactly just that. The A.I. if its morally upstanding, should build a network to exploit resources in a renewable way and distribute them as needed in human way, like if i need insulin then i get it and if you need more food since you are taller than me then you get it, and if it does it in a renewable way then there should be enough to fullfil requests like entertainment systems like tv and a pc able to run games, confier living conditions, custom clothing and extra decoration, and even swimming pools and ice cream for everyone!!! So that we pursue more human activities like art, science or just chill with people we like. I do believe that the fate of humanity is gonna be dependent on the fact that we build owr own A.I. overlord since our current human overlords are not gonna solve actual problems like global warming and world hunger.
You've seen the movies. You know it's inevitable.