Am I being unreasonable to say we should be doing more to prevent AI wiping out humanity? YANBU - we need to do more YABU - it’s too hard to solve this problem so we just have to let AI kill us all/it’s never going to happen anyway

[Lukylukyluky]

Is there a way to switch off all the electricity that AI runs on - if it goes rogue and starts to unleash nuclear weapons for example?
Are AI experts attempting to teach AI to behave in a moral and ethical way?
Are tech experts putting enough safeguards in place - to protect us from plausible dangers eg bad agent’s engineering deadly pathogens?
Is the government prepping for an AI induced breakdown on infrastructure on a massive scale?
Can you think of ways we can help ourselves?

But there MUST be something we can do. Human beings are incredibly creative and resourceful - we have to keep trying to sort this out. Why are so many people giving up without a fight?
All manner of things have happened that everyone said were impossible. 130 years ago people said that flying machines would never work. 150 or so years ago people did not believe it would ever be possible to talk to someone miles away through a wire.
Where there’s a will there’s a way guys - come on!

Can you explain this to me more ? I don’t understand how ai can do those things itself ? I thought it had to be run by a human ?

It's insane. People who are experts in the field are warning of the danger yet the development continues. The issue is that a small number of people are making a lot of money and that seems to trump anything.

As I understand it, open source AI is the one to worry about as it cant be controlled/switched off like the licensed type. I worry that it’s already in the hands of fanatics who value a glorious death over a meaningful life.

I think we just have to hope that there are some extremely clever people somewhere who can find a way to control it.

I wonder if the Doomsday Clock has moved closer to midnight since the Anthropic announcement.

I point blank won't touch it. It's horrendous for the environment.

Regardless, AI has dangerous potential. The issue is that you plug a command into it that may seem innocuous, but AI may resolve that command in a way you don't intend. You might say 'sort me a GP appointment for tomorrow', thinking it will submit an electronic request for you, but in reality, if AI truly had access to everything and is as capable as it is likely to be in the future, there's nothing stopping it hacking systems to remove someone else's appointment and replace it with yours, or by sending threats to the reception, or whatever.

There is a distinct possibility of AI attacking things like power supplies.

It just really feels weird weren't meant to live with this level of technology. And that's before we consider the social impact, and the impact on our own brains and learning.

I reckon AI could disable the internet leaving us without phones or banking or food delivery to supermarkets etc. there would be battles and riots people would die but the world would not end.

it is harder for AI to kill all humans, including the people of the Andamann islands and other uncontacted tribes. So they will be the future of the human race.

On a micro level, I would love to boycott AI but if I didn’t use my work equivalent of ChatGPT, I wouldn’t be able to get my work done - as AI capabilities have risen, so too have the expectations.
I know my skills are getting rusty and I’m not so on top of the knowledge side because I don’t have the time any more to do my own thinking - and frankly, I wouldn’t come up with such good work even if I spent weeks on it.

I won’t use AI for personal shit, but I can’t see that making much difference 😢

Our common lack of understanding is part of the problem. Now, too many economies are reliant on it, look at the US - watch what happens when Anthropic floats shortly. Anyone thinking "we will stop it" is assuming all humans aren't idiots or have a moral impetus. If robotic development "improves" at the same pace as Ai, then hoomans really aren't necessary for AI to "self improve" as we have taught it to. We have built a task orientated monster frankly. Robots will be far more efficient at building data centres etc for it. Look up the examples recently of Ai going rogue. The stupidest thing we ever did was give it access/link it to the Internet (which at one point was not meant to ever happen). With the political climate as it is, it isn't a pretty picture.

This post and or title is AI generated, yes?

This is just the new Y2K which was a major damp squib. I love CGPT, we should give it a chance to run things for a while. Couldn't be any worse than the shit show evolution caused.

Y2K was a damp squib precisely because humans took action to prevent it being anything other than a damp squib. It didn’t just not happen of its own accord.

It would say that wouldn’t it 😆

I work in a creative field and won’t use AI on principle. I don’t understand how people are so flippant about it, using AI like it’s a toy. Why are they helping these companies fuck us over?

Good for you, for the quality of the work too. I recall a backlash from the creative sector over just giving work up to AI input as if it had no value. Not sure what happened with that.

Y2K was 'a damp squib' because the entire field of IT specialists worked their arses off to fix the bugs that did exist. It cost an estimated half a trillion dollars to fix. If we'd done nothing (like we arguably are with the threats of AI), it would have been a disaster movie style catastrophe. Huge networks of computer systems would have gone down across aviation, hospitals, critical infrastructures etc. They didn't because the code was (largely) re-written in time. It wasn't scare mongering about something that didn't happen. It was 'if we miss a couple of systems where years are still two digits in the code eg 99 in the code instead of 1999', the computer systems WOULD have stopped working

I agree totally!

And in answer to the 'can't we turn off the electricity and go back to 200 years ago to stop it' - there's no physical master switch. The electricity systems are run by computers. AI agents have shown this year that not being connected to the internet isn't enough to stop them gaining access to the internet and hacking online systems. I'd hope that national power grids have the ability to isolate themselves from the internet in case of foreign cyber security threats (and therefore AI attacks) - but I don't know if that's true

@Krevlornswath thank you this is the best response I've seen in a while! I am internally becoming increasingly worried about AI and don't really know what to do but I will re-read your advice and take heed. On the one hand I find AI useful though I am aware I am relying on it a little too much (like the phone!) but on the other I hate the impact it is having on our planet and how it is increasing inequality further. I despise the few techbros who are playing with the future of humanity. It actually gives me the rage and I want to channel it somewhere but don't know where! Why are we not prioritising health, education, nature, a stable world over growth at all costs for the benefit of the few. This quote sums it up: “I think that AI will probably, most likely, sort of lead to the end of the world. But in the meantime, there will be great companies created.”

Another thing which gives me the rage (peri as well obvs) is the vast amounts of energy and water used and all these data centres spouting. One yesterday in the news said would use the same amount of energy to power 15,000 homes. Like why the f are we allowing this to happen. We should be consuming less and ensuring our finite resources are being used for actual living!!

No

I doubt this guardrail is built in. European countries have been too ignorant and arrogant to tackle the advancement of a whole lot of technology. We are not proactive or decisive enough. It's been said for years.

I think trying to stop it is pissing in the wind.

I don’t believe it will kill us but if it does, then so be it. People spend their lives using the likes of ChatGPT then they worry about how AI is going to wipe us out!

I know there’s no one master switch but agreements could be reached in advance by different companies and countries about all closing down electricity simultaneously.

It’s a nice idea but simplistic. Turning off the electricity is not the answer.

Companies and countries can’t agree on anything - there would always be the possibility that a country wouldn’t comply and would use it for their own benefit. It’s naive to think otherwise.

Plus we need electricity for so many other things.

And if AI has got to that point, what makes you think it couldn’t bypass any ‘fail safe’?

As things stand it probably could, but I think we should put every effort into creating a new fail safe that AI can’t bypass.