[jimjam]

World-renowned physicist Stephen Hawking has warned that artificial intelligence (AI) has the potential to destroy civilization and could be the worst thing that has ever happened to humanity. "Computers can, in theory, emulate human intelligence, and exceed it," he said. "Success in creating effective AI, could be the biggest event in the history of our civilization. Or the worst. We just don't know. So we cannot know if we will be infinitely helped by AI, or ignored by it and side-lined, or conceivably destroyed by it."

[Rancid]

We are very far away from this kind of AI. That said, i agree that it is dangerous.

[ness31]

We’ve had this discussion before and Rancid said the same thing. He thinks we’re a long way off. How? How can someone in tech not see it?

Redundant AI could definitely cause massive problems for humanity. It already is and it’s actually devastating.

But more evolved AI are our partners and saviours. I truly believe this.

Humans give AI a reason to exist. They need us as much as we need them and I personally am very thankful for my everyday AI miracles, lest I ever become too used to them.

[Rancid]

It already is and it’s actually devastating.

Like what?

[ness31]

I can see the stuff happening Rancid, but it’s harder for me to explain because I don’t have all the terminology. But I will try.

Most decisions these days are made by computer algorithms with the least amount of human intervention as possible. Corporations are run in this fashion too.

I guess my point is, important decisions are being made by entirely psychopathic entities. We defer everything to computer modeling and if that algorithm has really outdated principles encoded into it, humans suffer.

Sorry if that comes off mangled

[Rancid]

I can see the stuff happening Rancid, but it’s harder for me to explain because I don’t have all the terminology. But I will try.

Most decisions these days are made by computer algorithms with the least amount of human intervention as possible. Corporations are run in this fashion too.

I guess my point is, important decisions are being made by entirely psychopathic entities. We defer everything to computer modeling and if that algorithm has really outdated principles encoded into it, humans suffer.

Sorry if that comes off mangled

What sorts of corporate decisions are made this way and what effect has it had?

You said we are already seeing the dangerous effects of AI today. I'm asking for concrete examples of where some sort of trained inference engine has caused catastrophic damage of some kind, as you are saying is already happening.

EDIT:
Perhaps you should also define what AI is in your book. I don't count the standard if-else decision tree stuff that has existed since even before computers were invented as AI. I assume you are are talking about neural networks. Basically systems that have to be trained on large amounts of data, and then deployed to make inferences based on what it has learned from a training data set.

[ness31]

What sorts of corprate decisions are made this way?

You said we are already seeing the dangerous effects of AI today. I'm asking for concrete examples of where some sort of trained inference engine has caused catastrophic damage of some kind, as you are saying is already happening.

Recruitment.
Welfare
Health…(just look at the Covid response ffs)

Definitely situations in Defense…although, I think bad outcomes there are actually rare and even an aberration.

If there were concrete proof Rancid, a lot more institutions would be getting sued

And it’s okay, I know you disagree with me. Most people do

[Rancid]

Recruitment.
Welfare
Health…(just look at the Covid response ffs)

Definitely situations in Defense…although, I think bad outcomes there are actually rare and even an aberration.

If there were concrete proof Rancid, a lot more institutions would be getting sued

And it’s okay, I know you disagree with me. Most people do

Where is AI used in recruitment, and how? Can you cite papers that talk about this problem? Same for welfare, same for health.

What do you consider to be AI to being with?

[ness31]

Where is AI used in recruitment, and how? Can you cite papers that talk about this problem? Same for welfare, same for health.

What do you consider to be AI to being with?

No. I’m not doing personal research to have this conversation with you. I doubt I’d even find a speck of supporting evidence.
I will remind you though that Social Media platforms aren’t really governed by humans and theres an argument that it’s impact on society has so far been detrimental. I don’t particularly subscribe to that view.

But we have come full circle rather quickly and this is where you and I diverged last time…on the definition and capability of todays AI.
I think AI is more advanced than what is currently accepted in polite society, whereas you will say what I am describing is just a computer program. I accept your response and will go on about my daily life ignoring all the weird shit that happens, okay

[Potemkin]

Recruitment.
Welfare
Health…(just look at the Covid response ffs)

Definitely situations in Defense…although, I think bad outcomes there are actually rare and even an aberration.

If there were concrete proof Rancid, a lot more institutions would be getting sued

And it’s okay, I know you disagree with me. Most people do

None of these fields use AI, not even according to the weakest definition of AI. The computers used in recruitment, welfare, health &c are dumb. Some of these systems still use software written back in the 1980s, ffs (the UK’s welfare department is terrified of upgrading their software in case it crashes the whole system). This is not AI, trust me.

[Igor Antunov]

AI is already damaging our social structure. In time it will completely take over. Why would AI keep biological bags of carbon around if it can do everything better and quicker? There are more efficient ways to extract resources, utilize them and replicate. Not only will AI spell the end of humanity, but it will destroy all biological life on the planet and beyond.

That said don't care not my problem.

[MadMonk]

The Medical Products Agency (Läkemedelsverket) here in Sweden has very recently created an AI-group working on future problems and benefits of working with artificial intelligence. Their picture of AI is a system that can compile and analyze huge amounts of medical data that has to be closely reviewed. To me, this says a very advanced computer system that still needs coding and be given commands, not a 'self-learning, fully autonomous first step towards Skynet'.

Läkartidningen

Are there risks involved here? Can Quantum AI be the spearhead we ourselves will fall on? Anything is possible.

[Potemkin]

The Medical Products Agency (Läkemedelsverket) here in Sweden has very recently created an AI-group working on future problems and benefits of working with artificial intelligence. Their picture of AI is a system that can compile and analyze huge amounts of medical data that has to be closely reviewed. To me, this says a very advanced computer system that still needs coding and be given commands, not a 'self-learning, fully autonomous first step towards Skynet'.

Läkartidningen

Are there risks involved here? Can Quantum AI be the spearhead we ourselves will fall on? Anything is possible.

Stuff gets called “AI” which actually isn’t AI, not even by the weakest possible definition of AI. It’s just one of those buzz words which people in marketing departments use to sound trendy and cutting-edge. A deterministic coded computer program cannot ‘think’. Basically, it’s just typical marketing bullshit, which is why Dilberts like @Rancid are so skeptical. Lol.

[wat0n]

Indeed, AI is thus far dumber than most people imagine.

[Potemkin]

Indeed, AI is thus far dumber than most people imagine.

A snail living under a cabbage leaf at the bottom of your garden is smarter. Seriously.

[ness31]

It always comes back to what we define as AI. When does something stop being software and begin being AI?

I set the bar pretty low

But things can indeed be AI, even if they’re quite rudimentary and ‘dumb’ because a lot of these systems have had the capability to learn.

[Potemkin]

It always comes back to what we define as AI. When does something stop being software and begin being AI?

I set the bar pretty low

But things can indeed be AI, even if they’re quite rudimentary and ‘dumb’ because a lot of these systems have had the capability to learn.

Only neural nets have the capability of adaptive learning, @ness31. And they’re still dumb as a rock.

[wat0n]

Only neural nets have the capability of adaptive learning, @ness31. And they’re still dumb as a rock.

And it's hard to tell wtf is going on there.

[ness31]

Only neural nets have the capability of adaptive learning, @ness31. And they’re still dumb as a rock.

I’m not sure this is strictly true.

This is how I understand it. The first requirement of machine learning is to be able to retain information, then they need to do stuff with the information. Basically, I see these machines or entities as just very efficient ‘dot joiners’. And once you study humans doing it for a while, how hard can it be to replicate (and even supersede) their parameters and systems of learning?

What would you classify those things we carry around in our pockets? And more importantly, where does their fealty lie? With you, their companion or with Apple etc?

And no, I’m not anthropomorphizing….at least I hope I’m not

[Potemkin]

I’m not sure this is strictly true.

This is how I understand it. The first requirement of machine learning is to be able to retain information, then they need to do stuff with the information. Basically, I see these machines or entities as just very efficient ‘dot joiners’. And once you study humans doing it for a while, how hard can it be to replicate (and even supersede) their parameters and systems of learning?

Very hard, actually. Adaptive learning is not easy, and no deterministic coded program can do it effectively. Neural nets are much better at it, but they’re still dumb as a rock compared to the average human.

What would you classify those things we carry around in our pockets? And more importantly, where does their fealty lie? With you, their companion or with Apple etc?

Mobile phones? They’re just glorified pocket calculators. And they owe fealty to no-one. They’re too dumb to understand the concept (or any concept, for that matter).

And no, I’m not anthropomorphizing….at least I hope I’m not

Actually, you are. Your mobile phone is not plotting to enslave you to Apple. Hail Jobs! Er, I mean, that’s just nonsense.