Well, you haven't shared where you're getting your data. I'm just going from the various mainstream news stories, ignoring the "OMG abortion" fringe sources. Sure, the mainstream news sometimes fabricates stories entirely from nothing, but it seems reasonable to start from the assumption that the stories aren't entirely fiction.
BTW, most of Europe has stricter abortion laws than the US, so that's hardly surprising. Does the government make sure everyone knows screening tests are available, or is that part made up? And people have time to do the screening and act on the results within the law?
Ha! Yup, pretty much the same problem. What always amazed me was how easy those tests generally are, but kids are so bad at learning/generalizing because they're only taught the test that the teacher has no time to do anything but teach the test. Nasty feedback loop, there.
The basics of how individual neurons work is fairly well understood, and there's been remarkable progress in the past couple decades in understanding how simple neural systems work. Researchers are now doing stuff like modeling the simplest brains down to each neuron, and testing the model against the source (with reasonable success). It's enough to confirm we're not totally off-base.
Oh, I think it's more like 20 now. Neural networks were just too constrained by raw compute power until recently. That's why the earliest commercial applications were mainframe-sized voice recognition, a task which still mostly requires the mothership to be any good at. Suddenly there's a cloud's worth of compute, and lots of commercial funding form companies that expect results, so there's been serious acceleration the the field. Plus the well-funded war between reCaptcha and spammers has made massive progress in handwriting recognition accuracy, so we've seen some real results.
This is a constant real-world problem with most of the AU approaches - if you make them too big relative to the problem, they'll just "memorize" the training data. That is, they'll over-optimize on the specifics of the training data and not generalize well at all outside of it.
Neural nets do actually work in the way the neurons work, at least abstractly. Sure, the implementation is a bit different, as it's all just a bunch of matrix math and normalization, rather than an analog "wire logic" network, but the computational result is similar. It's more a matter of scale (AI neural nets are quite small) and refinement (who knows how many layers of optimizing-how-to-optimize even a simple animal brain has).
Neural Networks are nothing more than an approximation of a math function. Mathematically they are analogous to spline interpolation or Taylor series expansion. The only difference is that splines and Taylor series have a well known method to figure out the unknown parameters. Neural nets are just trained by finding a minimum in parameter space. Like splines and Taylor series, these donâ(TM)t work outsides of their bounds.
This is literally nothing intelligent put them.
While all of that is true, and worse (they tend to optimize to the first local minimum they stumble upon, which might be a poor choice), don't exaggerate the difference between that and how the brains of simple animals work. If we can model a space as a set of objects, that's more than half the battle.
Your sarcasm is noted, but it's still a pretty cool field. As much as everyone pointlessly frets about self-aware "AI" taking over, it doesn't seem farfetced that we'll see a collection of mahcine learning bit achieve the intelligence of, say, a chicken in our lifetimes. Able to train and optimize from general sensory data, not carefully chosen examples with perfectly matched feedback.
There are about a dozen different approaches to machine learning, and the neural net approach is probably the oldest in terms of being useful for something. None of them are "smart": all they can do is optimize, mostly randomly, until they succeed.
Image recognition in particular is something that has proven hard for machine learning, perhaps because the categories are fuzzy, or perhaps because humans are so good at it and that's the bar for comparison.
The classic, textbook example is handwriting recognition (even the simplified version of just recognizing numbers). Neural nets in particular just aren't great at even that simple task, though they were "good enough" for a lot of early products. If you look at the actual details of how they match, it's not at all how you'd think - there's no system of finding lines and loops and then trying to identify glyphs - just a large set of seeming-random heatmaps that somehow give "good enough" results. As I understand it, the modern stuff (the result of an amazing arms race between spammers and Google recaptcha) isn't neural-net based at all, but a far more sophisticated mix of models and hand-tuning.
In any case, none of the "AI" approaches have any sort of general problem-solving intelligence. They're merely self-optimizing systems tuned to a specific task. Which, to be fair, is what a lot of our neural matter is - but not the important bits for actual awareness.
So, explain Scandinavia? That group of countries has done more than anywhere else in the world to eliminate any sort of gender-based discrimination in career choice, at every level from the earliest age. The result? Tech is ~95% male, nursing is ~95% female.
People want different things. Good for them! What a boring world it would be if we were all the same.
Ah, so we're both annoyed by claims that Iceland has "cured Down's Symdrome". So, sure, 1-2 Down's births a year in Iceland, mostly due to the test being imperfect. Pedantry FTW.
But the moral questions raised by a society choosing to abort (effectively) all fetuses with a particular undesirable quality: those questions remain. And never doubt: this will become a larger effect on the species over time, as the technology becomes commonplace for more and more conditions. Couple that with fallible human judgement about the long term, and it could get pretty ugly.
Don't you agree we should be addressing, not ignoring, the moral and practical considerations while it's still early days?
Which, while a silly example, tells us that if an AI determines "kill all the humans" is possible and achieves its goals, it will.
Machine learning just finds optimal solutions. If the inputs allow for it, those 'optimal' solutions could include some outcomes we don't like... and since nobody really knows the decision process and what rules it's made for itself, you'll simply never know until it's too late.
Sooner or later, one of these things is going to do something exceedingly dangerous and costly, and nobody will see it coming or know why it happened.
Machine leaning doesn't think outside the box. It always optimizes within the constraints you give it. It's not going to jump to "hey, baby, want to destroy all humans?" unless you deliberately put that option there. These systems do not and cannot have "general intelligence".
None of which has anything to do with machine intelligence, or "strong AI" as some like to call it - a machine with general intelligence, the ability to generalize not merely optimize. Let's not build those.
I fully expect any AI applied to the stock market would eventually conclude stock manipulation would maximise return, and start breaking the law.
That's of course been happening for at least a decade, maybe two. High velocity program trading is an entire ecosystem of machine-learning-tuned algorithms deceiving machine-learning-tuned algorithms. It's all trolls trolling trolls at this point.
That part where people claim Iceland has "Cured Down's syndrome" instead of saying "we use eugenics to avoid substandard people". I've read such newspaper articles.
Don't get me wrong here, I'm not sure whether the practice is moral, amoral, or immoral. I think that's a very hard question, and not one to be taken lightly. But ignoring the question by pretending it's not eugenics? I'm sure that's the wrong answer.
Trolley problems are a bit contrived. Let's just say I'm not sure what the morally correct answer is. Nor should you be, unless informed by religious faith.
I don't know whether abortion is murder - it certainly seems to be late in a pregnancy, and I'm pretty skeptical that it would be early in a pregnancy, but what do I know? What do you know? I'm 99% sure that souls aren't a thing, but would you fire a bullet in the air in a city if it had a 1% chance of killing someone? The actual odds are a lot less, but it's illegal everywhere civilized.
I'm nearly certain, however, that being a vegetarian on moral ground while being a pro-choice absolutist is blatant self-serving idiocy, and there are a lot of those running around.
Question your deeply-held beliefs from time to time, man, it's important.
As long as it's under the ~24 week mark, the mother should be able to abort for any reason she wants. It's less about the fetus, and more about forcing the mother to host and birth a child she may or may not want.
Why 24 weeks? Why not 2 years? What different about detecting downs syndrome in the fetus, and discovering severe autism in a toddler? Do you know, with moral and philosophical certitude, when a person becomes a moral entity? I don't see how you could, unless your certainty is a result of religious faith. Or just lack of thinking about the question has left you with an unquestioned belief.
What about Gay babies? What about babies with the wrong racial ancestry? What about babies with merely below average IQ? Autism? What about 2-year-olds?
Eugenics raises serious philosophical and moral questions, and to be dismissive of it, or pretend it's not eugenics, is the only answer I'm sure is wrong.
It's still a case of "curing" a genetic condition through eugenics. I'm sure sure whether it's evil, but it sure is creepy. We shouldn't pretend it's anything other than aborting babies with undesirable genetics.
Slashdot Mirror
How long have brains been around for? 40 years is nothing.
Well, you haven't shared where you're getting your data. I'm just going from the various mainstream news stories, ignoring the "OMG abortion" fringe sources. Sure, the mainstream news sometimes fabricates stories entirely from nothing, but it seems reasonable to start from the assumption that the stories aren't entirely fiction.
BTW, most of Europe has stricter abortion laws than the US, so that's hardly surprising. Does the government make sure everyone knows screening tests are available, or is that part made up? And people have time to do the screening and act on the results within the law?
You sure seem touchy about this.
Ha! Yup, pretty much the same problem. What always amazed me was how easy those tests generally are, but kids are so bad at learning/generalizing because they're only taught the test that the teacher has no time to do anything but teach the test. Nasty feedback loop, there.
The basics of how individual neurons work is fairly well understood, and there's been remarkable progress in the past couple decades in understanding how simple neural systems work. Researchers are now doing stuff like modeling the simplest brains down to each neuron, and testing the model against the source (with reasonable success). It's enough to confirm we're not totally off-base.
Computers that can cheaply, quickly do matrix math with thousands of rows are new.
Oh, I think it's more like 20 now. Neural networks were just too constrained by raw compute power until recently. That's why the earliest commercial applications were mainframe-sized voice recognition, a task which still mostly requires the mothership to be any good at. Suddenly there's a cloud's worth of compute, and lots of commercial funding form companies that expect results, so there's been serious acceleration the the field. Plus the well-funded war between reCaptcha and spammers has made massive progress in handwriting recognition accuracy, so we've seen some real results.
This is a constant real-world problem with most of the AU approaches - if you make them too big relative to the problem, they'll just "memorize" the training data. That is, they'll over-optimize on the specifics of the training data and not generalize well at all outside of it.
Neural nets do actually work in the way the neurons work, at least abstractly. Sure, the implementation is a bit different, as it's all just a bunch of matrix math and normalization, rather than an analog "wire logic" network, but the computational result is similar. It's more a matter of scale (AI neural nets are quite small) and refinement (who knows how many layers of optimizing-how-to-optimize even a simple animal brain has).
Neural Networks are nothing more than an approximation of a math function. Mathematically they are analogous to spline interpolation or Taylor series expansion. The only difference is that splines and Taylor series have a well known method to figure out the unknown parameters. Neural nets are just trained by finding a minimum in parameter space. Like splines and Taylor series, these donâ(TM)t work outsides of their bounds.
This is literally nothing intelligent put them.
While all of that is true, and worse (they tend to optimize to the first local minimum they stumble upon, which might be a poor choice), don't exaggerate the difference between that and how the brains of simple animals work. If we can model a space as a set of objects, that's more than half the battle.
Your sarcasm is noted, but it's still a pretty cool field. As much as everyone pointlessly frets about self-aware "AI" taking over, it doesn't seem farfetced that we'll see a collection of mahcine learning bit achieve the intelligence of, say, a chicken in our lifetimes. Able to train and optimize from general sensory data, not carefully chosen examples with perfectly matched feedback.
There are about a dozen different approaches to machine learning, and the neural net approach is probably the oldest in terms of being useful for something. None of them are "smart": all they can do is optimize, mostly randomly, until they succeed.
Image recognition in particular is something that has proven hard for machine learning, perhaps because the categories are fuzzy, or perhaps because humans are so good at it and that's the bar for comparison.
The classic, textbook example is handwriting recognition (even the simplified version of just recognizing numbers). Neural nets in particular just aren't great at even that simple task, though they were "good enough" for a lot of early products. If you look at the actual details of how they match, it's not at all how you'd think - there's no system of finding lines and loops and then trying to identify glyphs - just a large set of seeming-random heatmaps that somehow give "good enough" results. As I understand it, the modern stuff (the result of an amazing arms race between spammers and Google recaptcha) isn't neural-net based at all, but a far more sophisticated mix of models and hand-tuning.
In any case, none of the "AI" approaches have any sort of general problem-solving intelligence. They're merely self-optimizing systems tuned to a specific task. Which, to be fair, is what a lot of our neural matter is - but not the important bits for actual awareness.
I think you're just arguing with a trolling account.
So, explain Scandinavia? That group of countries has done more than anywhere else in the world to eliminate any sort of gender-based discrimination in career choice, at every level from the earliest age. The result? Tech is ~95% male, nursing is ~95% female.
People want different things. Good for them! What a boring world it would be if we were all the same.
Whenever they try that shit with me, I ask them to explain what happened in Rwanda.
Sadly the answer in South Africa is "watch this".
Ah, so we're both annoyed by claims that Iceland has "cured Down's Symdrome". So, sure, 1-2 Down's births a year in Iceland, mostly due to the test being imperfect. Pedantry FTW.
But the moral questions raised by a society choosing to abort (effectively) all fetuses with a particular undesirable quality: those questions remain. And never doubt: this will become a larger effect on the species over time, as the technology becomes commonplace for more and more conditions. Couple that with fallible human judgement about the long term, and it could get pretty ugly.
Don't you agree we should be addressing, not ignoring, the moral and practical considerations while it's still early days?
Atari 2600 games were written in assembly language, however.
Which, while a silly example, tells us that if an AI determines "kill all the humans" is possible and achieves its goals, it will.
Machine learning just finds optimal solutions. If the inputs allow for it, those 'optimal' solutions could include some outcomes we don't like ... and since nobody really knows the decision process and what rules it's made for itself, you'll simply never know until it's too late.
Sooner or later, one of these things is going to do something exceedingly dangerous and costly, and nobody will see it coming or know why it happened.
Machine leaning doesn't think outside the box. It always optimizes within the constraints you give it. It's not going to jump to "hey, baby, want to destroy all humans?" unless you deliberately put that option there. These systems do not and cannot have "general intelligence".
None of which has anything to do with machine intelligence, or "strong AI" as some like to call it - a machine with general intelligence, the ability to generalize not merely optimize. Let's not build those.
I fully expect any AI applied to the stock market would eventually conclude stock manipulation would maximise return, and start breaking the law.
That's of course been happening for at least a decade, maybe two. High velocity program trading is an entire ecosystem of machine-learning-tuned algorithms deceiving machine-learning-tuned algorithms. It's all trolls trolling trolls at this point.
"Assembly language." An assembler is like a compiler.
That part where people claim Iceland has "Cured Down's syndrome" instead of saying "we use eugenics to avoid substandard people". I've read such newspaper articles.
Don't get me wrong here, I'm not sure whether the practice is moral, amoral, or immoral. I think that's a very hard question, and not one to be taken lightly. But ignoring the question by pretending it's not eugenics? I'm sure that's the wrong answer.
Trolley problems are a bit contrived. Let's just say I'm not sure what the morally correct answer is. Nor should you be, unless informed by religious faith.
I don't know whether abortion is murder - it certainly seems to be late in a pregnancy, and I'm pretty skeptical that it would be early in a pregnancy, but what do I know? What do you know? I'm 99% sure that souls aren't a thing, but would you fire a bullet in the air in a city if it had a 1% chance of killing someone? The actual odds are a lot less, but it's illegal everywhere civilized.
I'm nearly certain, however, that being a vegetarian on moral ground while being a pro-choice absolutist is blatant self-serving idiocy, and there are a lot of those running around.
Question your deeply-held beliefs from time to time, man, it's important.
As long as it's under the ~24 week mark, the mother should be able to abort for any reason she wants. It's less about the fetus, and more about forcing the mother to host and birth a child she may or may not want.
Why 24 weeks? Why not 2 years? What different about detecting downs syndrome in the fetus, and discovering severe autism in a toddler? Do you know, with moral and philosophical certitude, when a person becomes a moral entity? I don't see how you could, unless your certainty is a result of religious faith. Or just lack of thinking about the question has left you with an unquestioned belief.
So you're going to adopt all the unwanted babies? If not, you're hand-wringing is just shallow concern-trolling.
Perhaps parents have a moral duty to care for their offspring, wanted or not?
Perhaps the state has a duty?
Perhaps we just kill anyone overly inconvenient?
Do you think these are easy questions?
Why is it my business the reason a woman kills her toddler? Same reason. It's the same moral question regardless of the age of the child.
Again, the only answer here that's certainly wrong is being dismissive of the question.
>Iceland needs moar retard babbies!
You Amerikuks are hilarious.
What about Gay babies? What about babies with the wrong racial ancestry? What about babies with merely below average IQ? Autism? What about 2-year-olds?
Eugenics raises serious philosophical and moral questions, and to be dismissive of it, or pretend it's not eugenics, is the only answer I'm sure is wrong.
It's still a case of "curing" a genetic condition through eugenics. I'm sure sure whether it's evil, but it sure is creepy. We shouldn't pretend it's anything other than aborting babies with undesirable genetics.