SoylentNews is people
SoylentNews
The hysteria about the future of artificial intelligence (AI) is everywhere. There seems to be no shortage of sensationalist news about how AI could cure diseases, accelerate human innovation and improve human creativity. Just looking at the media headlines, you might think that we are already living in a future where AI has infiltrated every aspect of society.
While it is undeniable that AI has opened up a wealth of promising opportunities, it has also led to the emergence of a mindset that can be best described as "AI solutionism". This is the philosophy that, given enough data, machine learning algorithms can solve all of humanity's problems.
But there's a big problem with this idea. Instead of supporting AI progress, it actually jeopardises the value of machine intelligence by disregarding important AI safety principles and setting unrealistic expectations about what AI can really do for humanity.
In only a few years, the pendulum has swung from the dystopian notion that AI will destroy humanity to the utopian belief that our algorithmic saviour is here.
[...] Examples demonstrate that there is no AI solution for everything. Using AI simply for the sake of AI may not always be productive or useful. Not every problem is best addressed by applying machine intelligence to it. This is the crucial lesson for everyone aiming to boost investments in national AI programmes: all solutions come with a cost and not everything that can be automated should be.
What is your take on this? Do you think AI (as currently defined), can solve any of the problems, man-made and otherwise, of this world?
(Score: 3, Insightful) by Gaaark on Tuesday May 29 2018, @03:17AM (21 children)
Unless an ai can control people, it won't change a thing:
Example...say the solution is "no one can have more than $100,000 in total assets, and the world will be 'its best'.
Will the rich give up what they have to solve the world's problems?
The solution is "no one can eat ANY meat ever again".
The solution is "no one can own/possess/manufacture any kind of weapon".
"No one can ever have sexual relations ever again (whether it be by proper definition or Clinton definition)".
AI will not change people and only people can save the world.
"I'm a man. But I can change. If I have to. I guess."
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by mhajicek on Tuesday May 29 2018, @03:35AM (2 children)
The better AI will tell you how to convince or manipulate everyone to follow the plan. Unfortunately at present the plan is to funnel all resources into a few pockets.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 1) by Ethanol-fueled on Tuesday May 29 2018, @05:36AM (1 child)
We are already seeing the "better AI," and it's not doing much but pissing people off and pushing Google search users over to Bing search, and pushing YouTube users over to HookTube. Most of us are too smart to be fooled by AI tricks. That why I have faith in humanity over deep-state technology, it is very obviously not yet good enough to beat us.
(Score: 2) by mhajicek on Tuesday May 29 2018, @05:59AM
Chess, Go, StarCraft, salesmanship. Seems like a logical progression.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 1) by khallow on Tuesday May 29 2018, @03:54AM (9 children)
If that's the solution, drop the problem and find a better one. Solving ill-posed or ill-conceived problems doesn't do much for us and may actually cause considerable harm.
(Score: 2) by acid andy on Tuesday May 29 2018, @05:38PM (6 children)
Funny; I was thinking almost exactly the same thing but about the sexual relations example instead. If it specified procreation instead, then fine.
The $100,000 one didn't sound like such a bad idea, if there was a way to index link it to the Earth's resources (all elements and organisms, not just gold).
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 1) by khallow on Tuesday May 29 2018, @10:20PM (5 children)
So you want to incentivize humanity to take the Earth completely apart for its resources? Because there's a vast difference in that number between an intact Earth and one where the resources of Earth have been to the last pebble distributed to an entire Solar System civilization.
And if no one is allowed to have more than $100k, then where's the incentive to grow the pie? To the contrary, there is no use for such a constraint. We are not wealthier or better off because Bill Gates is worth $100k instead of $100 billion. And without massive incentives to make the world better in massive ways, who is left to do this stuff. You've just pushed a bunch of power into the lap of government without any corresponding increase in how much they'll care.
(Score: 2) by acid andy on Wednesday May 30 2018, @02:04AM (4 children)
Yeah, I did worry a bit about a gold rush for every resource, but the idea was to set the wealth cap at a low enough level that it actually limits how much of the Earth's resources can be plundered, rather than accelerating their extraction.
It certainly looks like it would have some problems. Presumably Bill would have to start spending his wealth before he was allowed to earn anymore. The trouble is, he'd only be allowed to spend it on people that hadn't yet got their full $100k, if there were any left. Hmm, if they weren't, how could anyone purchase anything? They could have children in order to palm some of the money off to them. Ah but Gaark's AI said that wasn't allowed!
Hey, Gaark, what happens when everyone on the planet has $100k in assets? Are they allowed to burn the assets? I suppose it would be more logical to burn assets of the person you want to purchase from, rather than your own. Yeah, this system would cause a global crime epidemic. People could always trade assets of equal value, but no-one would be allowed to make a cent of profit once they were at the cap.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 2, Insightful) by khallow on Wednesday May 30 2018, @04:10AM
What exactly is wrong with plundering resources? They're not doing anyone any good sitting in the ground. And the vast majority are so plentiful and easy to come by (even with that environmental footprint) that it doesn't make sense to keep it in the ground. Finally, we have over seven billion people to care for and improve the lives of. Artificial restrictions on resource consumption run contrary to what has worked amazingly well so far.
Let's look at that argument in more detail. First, resource extraction just isn't that bad environmentally outside of agriculture. The environmental destruction can be quite severe, but it's just that area that is affected. And we've developed a bunch of technologies and strategies for limiting our impact of resource extraction.
The real pollution problems happen after its dug out of the ground. But even that is manageable. It's a solved problem in the developed world.
And what's the point of saving in the ground merely so that another generation can also save it in the ground so that another generation can... At some point, resources have to come up in order for them to have value to us. It is merely a matter of whether it's better to mine them as soon as convenient or let them wait for a future generation. Here, the obvious thing is that our technologies of extraction, recycling, and such improves over time. So future generations won't have the same resource headaches we do now. So why coddle them?
Agriculture is a bit different, but it has the virtue of being heavily renewable. With sparing use of natural gas-based fertilizers, one should be able to get it to keep going for another century. By then we'll probably have long closed the gap in renewable nitrogen and perhaps figured out how to recycle phosphorus and other macronutrients as well (depends on whether we need to, of course).
Moving on, why the obsession with resources anyway? They're not that scarce and their exploitation doesn't cause that much in the way of trouble (particularly in the developed world). Meanwhile we're doing amazing things with those resources, such as elevating the entirety of humanity out of poverty and developing a global, high tech society. Don't we have better things to do with our time?
(Score: 2) by Gaaark on Wednesday May 30 2018, @02:12PM (2 children)
See my other response about the example being pulled out of my dogs ass: the example has no real relevance to.... anything.
Just a silly example.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by acid andy on Wednesday May 30 2018, @03:35PM (1 child)
Sorry for spelling your name wrong, by the way. The "a"s just seemed to blur together. My eyesight probably isn't what it once was. I guess Copy & Paste is my friend in these circumstances. ;)
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 2) by Gaaark on Wednesday May 30 2018, @03:57PM
Just don't call me late for supper.
;)
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by Gaaark on Wednesday May 30 2018, @01:36PM (1 child)
Just an 'example' pulled out of my dogs ass...don't take it for anything 'real'.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 1) by khallow on Thursday May 31 2018, @12:49PM
(Score: 0) by Anonymous Coward on Tuesday May 29 2018, @03:56AM
Then we better not give any AI a lethal weapon, or control over life support systems. Because "control" and "kill" may just mean the same to some AI.
But AI can save the world from people.
(Score: 2) by shortscreen on Tuesday May 29 2018, @05:34AM
I think AI could be a tool that would provide a medical diagnosis, help generate "creative" works (procedurally generated, but with a procedure fancy enough to be called AI), or work on engineering problems.
As you say, it's unlikely to help with any problem where the difficulty lies with getting large numbers of people onboard.
(Score: 2) by Wootery on Wednesday May 30 2018, @10:38AM (5 children)
Strong claims, and zero supporting evidence...
And you're basing this on what? That you can't imagine an AI being superhumanly persuasive?
Your failure of imagination does not demonstrate impossibility.
If anything, the evidence is against you. This experiment [yudkowsky.net] suggests than for an AI to escape an 'airgapped' environment, it need only be as persuasive as a skilled human. (It's annoying that Yudkowsky refuses to disclose his persuasive techniques.)
(Score: 2) by Gaaark on Wednesday May 30 2018, @02:01PM (4 children)
Basically I said that under the heading of "Unless an ai can control people, it won't change a thing:"
When an AI CAN manipulate people, this all changes (basically saying people will fuck things up unless control over things is taken away from them).
No argument with you otherwise.
Can't wait for an intelligent sexy sex-bot, lol.
"That's it, baby... manipulate me!"
;)
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by Wootery on Wednesday May 30 2018, @02:23PM (1 child)
Ok, I get you now.
I'm not sure I agree though. If we have AI running the stock-markets, they will have a great deal of power, even if they have zero ability to communicate with us directly.
(Score: 2) by Gaaark on Wednesday May 30 2018, @02:44PM
We DO have 'artificial' intelligence running the stock market: they're called traders, lol.
But yeah, seriously, this is why i try to support things like mycroft.ai : open source and open ermmm 'idealogy'??? AI with a conscience??? cause yeah, AI in the hands of corporations could REEEEEEEALLLLLLLY FUCK US AAAAALLLLL!
AI COULD be really good, but with the way things are heading.....
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by Justin Case on Wednesday May 30 2018, @02:55PM (1 child)
But AI can control people. We've arrived at the point where AI can kill people -- at will, without human oversight. See for example https://en.wikipedia.org/wiki/Death_of_Elaine_Herzberg [wikipedia.org].
Perhaps at present no AI "understands" the power it has... do you want to gamble that such a day will never come?
(Score: 3, Insightful) by Gaaark on Wednesday May 30 2018, @04:11PM
Hooboy!
No, AI has nothing to do with this case:
Uber 'turned down' the AI to almost zero, in order to avoid false braking. If they'd turned the AI "up to 11", this could easily have been avoided and she would still be alive.
It was from the top down in Uber to decide these things and has NOTHING to do with the AI.
Uber fucked up big time and tuned the software (the AI) to a point that Uber liked. She was killed by Uber and Uber alone and Uber should be bankrupted for it.
The AI did what it was allowed to do and braked at the last second.
The AI was ONLY allowed to do what it was ALLOWED to do
Don't blame AI for Uber executives being assholes (and no, it is not real AI). Uber should be gone, should be bankrupt and Uber's top executives should be in jail at the least. Uber should have been shut down... what does that tell you about the cost of living in todays world:
BIG CORP kills a person... who cares.
A random psycho kills a person....O.M.G! THEY USED A GUN AND KILLED A PERSON! O.M.G! UBER USED A CAR AND KILLED A PERSON AND NO ONE CARES BUT GUN/NRA/PSYCHO!
Random psycho is jailed...Uber continues as normal.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---