I think the REAL issue is unintended consequences wrt strong AI. Something vague like ‘protect and serve humanity’ could have very bad consequences, as could vague touchy feely sentiments about altruism or doing good. I think the whole AI overlord wanting to take over has been vastly overblown, but if we fuck ourselves by being vague in what the strong AIs goals and methods are, or just don’t think through the ramifications before we unleash it, could be nasty.
If anyone is interested in a fairly long video (YouTube), here is part one of Michio Kaku’s video on what things will look like (in his opinion) by 2045…and it’s got some interesting scenes about what predictive AI (weak) is already doing today for law enforcement and in other areas.
So let’s clearly and distinctly stress the importance of eliminating human error.
Lol. I mean, I hope that’s something I’m meant to lol about.
Yup. All sorts of fine-sounding goals – reducing human suffering, eradicating HIV, lowering the crime rate, you name it – are problems that can be solved by simply wiping out mankind. Voter fraud? Gone. Income inequality? Gone. We can end racism and sexism and homelessness and poverty by just giving a badly-worded request to a literal-minded servant.
Well we just slide in a “make humans happy” rule and we’re golden. Sure, we’ll all be on an opiod drip locked in an isolation pod but hey it could be worse.
AI is certainly moving along well now, but I think the next BIG thing is (Wired Magazine link). Claims (in the link) are improved computing speeds of 100,000 to 100,000,000 times better than current computers. Because quantum computers can analyze many different combinations simultaneously, playing the game of Go will seem like child’s play. I can barely imagine what QC will do to AI with those computer speeds.
LOL…I just posted in another thread, mentioning Ayn Rand…and who’s the very next responder I see?
Are you following me?
Actually, part of the point of AlphaGo was sidestepping the straight computer muscle of computing speed paradigm. It learned what strategies/moves were and weren’t useful to hammer out. That’s what made it so different from Chess playing computers.
That reminds me of a joke I heard from a comedian who was doing a set the week after CFL lights hit the market.
‘they say these lights will last 10 years, how do they know that? They just invented them last thursday’.
I’m assuming we will reach a point where medical tech is advanced enough that we can predict how well we can delay or reverse the aging process.
100 years is a long time as far as medical progress goes. Dolly was 20 years ago, we are talking nearly 100 years from now.
We are already examining various ways to reverse or delay the aging process via reductionist biology. Eventually we will progress to systems biology, then to actually rewiring our biological systems to reverse aging. The tech will continue to grow and get better. There will be bugs but as time passes we will become better at identifying and stopping the bugs.
The overall progress of technology is more and more rapid growth, and exponential growth in our information gathering and processing abilities over time. Far more science and information is gathered from 2007-2017 than was developed and gathered from 1907-1917. The same trend will continue.
-
Consumer demand. People want AI and to live longer, healthier lives. There are multi-trillion dollar markets for advances like this (global health care spending is almost 7 trillion right now, and most of that is for diseases of aging). Aging is becoming a serious problem in China, so China has a strong incentive to invest in robotics, automation, AI and anti-aging to keep their society functioning. Other western nations are facing the same thing. The massive consumer demand will create an environment where public and private sector work together to meet the demand. I could see the governments of China, the EU & US funding anti aging research in partnership with large tech and medical companies.
-
No, if anything there are social costs. Older people grew up in a time where there were more prejudices. Imagine if people who were born in 1830 were still alive today. It would hold back racial progress, feminism, gay rights, secularism, etc.
-
Probably via universal basic income. The means of production will eventually be socialized, and everyone given a stipend.
-
I’m not sure. Thats a good question, what if a computer gives what seems to be a nonsensical answer, but it would be like a human telling a squirrel to ‘run’, because a car is coming and the squirrel isn’t intelligent enough to understand what is happening. No idea.
-
I’m assuming colonization, due to depletion of natural resources will be a major factor. I’m also guessing it may be possible to circumvent the speed of light someday via something like an alcubierre drive. Even if not, and even if we can only travel at 10% the speed of light, generation ships will eventually colonize other solar systems because there may not be enough natural resources in ours.
I mean, isn’t the obvious fix that you enable AIs to draft plans, and to implement detailed plans they’re given, but not to implement the plans they draft until a human gives them the go-ahead?
I think it’s naive to assume that social changes brought by the newest generations will always go in the direction that you would consider a progress. It might be that the next generation or the one after it will be enthralled by authoritarianism or nihilism, or worst…
Sure, that’s how it will start. But eventually the “plans” will be so complex that no human will be able to understand the reasons for them, or the outcomes. So we’ll just implement them anyway without understanding them, because the last 9,999,999 times we did that we got excellent results.
Yeah, ask Europeans circa 1910 how much social progress the next generation is going to make.
It really is crazy how much the people of the first half of the 20th Century thought they could remake society and human beings like a set of tinker toys.
For all our problems today at least we’ve given up on utopianism.
The concept of progress is utopian.
I don’t understand how AI will improve social progress. What mechanism will allow that to happen?
Crane
Well, take medical care. Today it’s very expensive to provide medical care to everyone. If you could just see a robot doctor instead it would be much cheaper. There we go, medical care for everyone, paid for by the spare change we dig out of the couch cushions.
Even without robot doctors, if you just had a system that could optimize costs and care, could model and project what would be needed where and shift assets about, keep detailed track of and make projections about patient data and needs compared and contrasted world wide with other data and systems and you could potentially have a system that could be orders of magnitude more efficient and effective at a fraction of the costs than our current system. Now project that on every other human activity and you get the feel for the potential for strong AI and how it could improve ‘social progress’, whatever that means.
That is an interesting POV, why would you say progress is utopian? Ever since the industrial revolution, humanity has been progressing in the right direction.
Better health
More education
More wealth
Less poverty
more political and civil freedom
Less oppression of minority groups
More ability to understand and control our environment
More ability to stay safe in a hostile universe
The trendline seems pretty clear to me. Why would progress be utopian?
The world today is better than in 1917, and 1917 was better than 1817. There are wars and political setbacks (communism and fascism) but the overall trajectory is clear.
No snark or sarcasm intended but, it is your lack of imagination (and everybody else’s too btw). I was a teenager in the '70s and, though everything today seems mundane & banal, much of it was completely unimagineable and, for the most part, unimagined, by mere mortals like me. Even some of the nomenclature doesn’t work. I took a few “computer science” courses in the mid-'70s because