I think you’re massively downplaying the importance of the internet. It’s affected a lot more than just commerce, it radically altered our mode of communication. Being able to communicate with anyone, anywhere, and at anytime, instantly is revolutionary. It’s aided revolutions abroad and transformed the world of social interaction and advertising at home.
Not to mention the fact that the 50 years between 1950-2000 produced much more than just “the internet”, little things like huge advancements in heart disease treatments, the global eradication of smallpox and other diseases, car designs that are many times safer, GPS, cell phones, satellites, the entire field of computer graphics, a much better understanding of the human brain, etc. etc. You are seriously underrating the medical and technological progress that has taken place in the recent past.
I agree with pretty much everything **Voyager[\b] has said.
Intelligence is not necessarily the inevitable output of high computer speed. We really don’t know how to make a comprehensive AI. Maybe all we’ll get for our amazing speed improvements is better graphics, higher-resolution interfaces, and more calculating power.
The hardware changes coming will be revolutionary enough. Really fast computers and the continued shrinking of hardware is going to enable amazing applications. We’re going to extend our senses, supplement our bodies, probably lengthen our lifespans, and see dramatic improvements in our quality of life. We’ll be controlling machines with our brains, possibly in intuitive non-language ways. It’s going to be very, very cool.
But a singularity requires a strong AI, and we simply don’t yet know how to make one of those, or even if we can.
I do agree that if we ever see a true machine intelligence, it will be evolved rather than programmed. But that suggests we won’t really know much about it - just like we know very little about how the human brain works to create self-awareness and general intelligence. Its behavior won’t necessarily be predictable.
My thoughts on the Singularity is that it is, if it ever happens, a very, very long way off.
The reason is pretty simple: the phrase ‘What If’.
It is a simple little phrase yet it covers something that computers cannot really do and I don’t think that they ever will be able to unless there is a huge shift in the underlying technology.
Computers are binary as everyone knows and it seems to me that binary cannot represent concepts in a way that we humans can. Binary is either 1 or 0, on or off.
I believe that serious creative thinking goes beyond a simple 1 or 0, on or off, kind of thing. Serious creative thinking is ‘What If’ based, or at least it seems to me. When I say What If based, I don’t mean a simple choice, as in What If option a) or What If option b). I mean it in a “What If light bends in gravatational fields. What would that mean?” sense.
The What If thinking isn’t binary and it isn’t always logical. What If thinking is taking a leap, going past what is already known and coming up with a new way of looking at or thinking about a problem. And What If thinking seems to be, to me at least, a non-logical process. And What If thinking seems to use that thing we call creativity, though I don’t think anyone has a really good understanding of what that really is or how it really works in our brains. And I don’t know if we will every really understand it.
Now, I am sure Kurzweil would say that creativity is just a biological process and can be replicated or simulated in silicon. I don’t think I buy that. I don’t know how creativity happens but, when I am at my most creative, it isn’t a logical process. The step from “I don’t know” to “What If” doesn’t take a logical path a reasonable amount of the time when I solve tricky problems. The end result may be logical as hell, but the jump to get there doesn’t come from a logical thought process. Or at least it doesn’t seem to come from a logical process.
I like to use a line I stole from Feynman sometimes, ‘That’s it. Now I understand everything!’. I’ll be working on or thinking about a problem and someone will say something unrelated to the problem. But what they say lights off an idea and gives me a new insight to the problem and how to solve it.
I don’t think adding more transistors will ever allow computers to make those kinds of jumps.
Now, there are a whole shitload of problems that can be tackled by brute force.
As I understand it, the original concept of the Singularity arose, not from consideration of specific technologies like AI and nanotech, but from graphing the rate of technological progress over the last few centuries. The graph shows a curve that begins moving upward at a very low rate but over time that rate increases, and somewhere around 2030-2050 the line becomes almost vertical … the technological singularity, where technology changes so fast it becomes unpredictable.
AIs are just considered the most likely cause of a technological singularity. A s I understand it, the theory of the singularity was just some scientists trying to figure out the meaning of a graph. So, not quite the quasi-religious thing it’s been described as by some.
Yes, but you can make the same argument for any exponential growth curve. That was the logic that led to the belief that we’d be out of food by now and heavily over-populated. It’s also the logic used to ‘predict’ that we would have been traveling at significant fractions of the speed of light by now and flying around in supersonic airliners.
The problem with extracting a curve into the future as a prediction is that you never know if you’re near an inflection point. As they say, past performance is no guarantee of future returns.
That’s not even a very good example of how complex small talk is, so it falls kind of flat. Think about the comments that we make all the time that propositionally are simple, but still completely beyond the grasp of something like Watson because of context and affect, and yet which are utterances which can determine things of great import, such as whether someone gets a job, or whether to avoid a possibly threatening person.
Computers are capable of functioning in ways far more complex than just binary. Really, this is just like saying “humans are just made of atoms, and atoms can’t think”.
I have a feeling you might be talking about determinism though, but again, so what?
I’d bet against that, if there was any way to judge objectively (not to mention collecting on my win at age 106).
You beat me to it.
I almost agree. I do agree but only up to say 1990.
I remember, in the late 80’s talking with my grandfather and asking him about the first time he saw an automobile, when he was about 10. I surmised that his life span had seen the most dramatic changes in the structure of society, and wondered whether mine would see that much or more. At that time, I considered the automobile, telephone, and TV to have had the biggest impacts on how we lived and the structure of society. Computers were huge, but up to that point, had changed the way we worked but not lived so much (with one big exception: credit cards). Life here in the South must have been miserable before air conditioning!
But then came the web with the latest communications revolution, which I think is as dramatic in terms of how we live our lives as those earlier ones. The way we get anywhere is different, the way we plan any excursion is different, the way we keep in touch is different (mobile phones etc). When’s the last time you bought a new map?
Of course, as mentioned above, medical technology has accelerated. The difference there is merely that lots of people who would have died didn’t. That doesn’t change society so much, but it certainly touches lives.
I call BS. I don’t think we know nearly enough about qualia or experiences and how they’re formed to answer this. All we can say is that enjoyment correlates to releases of neurotransmitters. But I agree with your point that if a biological machine can do it, then an electronic machine can as well.
Regarding the original question, Kurz (as we keyboard players call him and his products) knows more than the vast majority of us about AI. Not only is he frickin brilliant, he’s made it his life’s study – both human intelligence and artificial. I thoroughly enjoyed his “Creating a Brain”, which gives a pretty good impression of how much we know and how much more we don’t.
Still, I think he’s being overly optimistic. On the other hand, I do believe that before long, things will start changing faster than many of us can manage to keep up (Tofler’s point in Future Shock, which I have not yet read and may be out of date, but I find the premise interesting). The fact that many of us won’t keep up will have an impact on how (and how fast) things change, but it won’t stop the changes.
I shudder to think how different the world might be in 50 years. I sure wish I could get a glimpse (for example, sample ordinary TV shows from various points in the future.) Oh well; I’ll have to settle for however long I live. I don’t expect to see a singularity in my lifetime (say, 30 more years, with a bit of luck.)
Computers don’t have to be binary or logical, oddly enough. That’s true without changing the hardware, too. We have the math to represent shades of gray (logically) and we can implement that on a digital computer.
There’s a lot of assumption that the singularity AI has to be based on how humans think. From a logical standpoint, a computer is already a better thinker than our emotional, crying husks of meat. I think that if we can figure out how to build experience into the logical processing, we’ll have pretty much made something smarter than humans.
Did it ask you to press F1 because it detected no keyboard?
The problem with this line of reasoning is that it fails to take into account how evolution changes things fundamentally. Yes, “Vehicles” made a big difference in interstate transport. But the original 18-wheel transport vehicle was very, very different than what we have, today. Today’s over the road trucking industry is even very different than what we had in the 90s.
The evolution of every aspect of trucking (not just the vehicles) is what gave us what we have today. Inventory control dropping a pick list to a warehouse operator to ready a shipment automatically speeds up load time. And, for stuff like grocery stores, you’ll pack 5 stores into a single box trailer and have them pulled off in the order the truck arrives at each store. Once that store is dropped, pull away from the dock and go to the next store. No need to count out what’s being delivered (“Let’s see, your store gets 12 apples, and 44 bananas.” etc)
On the trucks themselves, technologies all over a heavy truck tells the engine how much fuel it needs to use based on load and speed, slowing the burn of fuel if you aren’t hauling vs hauling a light load vs hauling a full 80,000 pounds. The transmission will report to the ECM that the fluid is getting too thick and it’s time for a change and so on.
The reason you can point to the trucking industry and make that statement, today, is because we have spent 70 years in this country perfecting every aspect of supply chain management. Not just that 18 wheelers showed up in the 40s and were an answer to everything concerning shipping. They were better at the time…and then we made them betterer.
Yes, we needed the original 18 wheeler to make interstate transport what it is. But we also needed efficient distribution, better vehicles, and better recipients to make it happen, also.
Every invention you can point to historically that was important can be shown to have a significant evolution to get to the point that it’s considered important or indispensable, today.
I agree, and the trucks of 50 years from now will be very different than today’s trucks. They will evolve: they may be driverless, they may be loaded by robots, etc…
But all these changes will be just that–gradual changes. They will not bring us anywhere nearer to the singularity.
That’s why I said in post #12, “life ain’t gonna change much in the next 50 years.”
Kurzweil is not going to see his fantasy come true. He’s going to die just like you and me, at age 85 or 90. Yes, the hearse may drive itself to the cemetery, and the clergyman may speak to the mourners remotely (thru google glass?), while robots fill the shovels with dirt to bury the casket----But life itself won’t be that much different in 2050 than in 2000.
We’ll be using the same toilets , the same refrigerators , shopping at the same supermarkets, watching the same TV shows.
We’ll still be going to work in an office, and working for a real, human boss, not some kind of mythical AI robot with superhuman brainpower .
I agree, the graph curve does not determine future history, it merely points to one logical extrapolation of present history. If the data points start changing, the curve will have to change with it. I was just pointing out that the origin of the technological singularity is logical in nature, not mystical, even if some have attached mystical significance to it.
I think those who hold that the evidence points to a technological singularity occurring circa 2040 would say, “Well you can say curve might change as we approach the Singularity, but the fact is it has NOT changed over the last few centuries and decades, it has held remarkably steadily. If you wish to say the Singularity will not happen, you have to come up with an explanation why the curve should change.”
As has been pointed out, progress on AI has not been encouraging in terms of a 2040 technological singularity, but breakthroughs do happen, and with the case of a transhuman-level AI, it only has to happen once … it’s not like you have to have mass-market AIs for it to have a profound effect on human society. So I sort of have a sceptical/open-minded attitude … prospects for the technological singularity occurring in the next few decades look dim, but not impossible. Breakthroughs happens.
If an AI with superhuman brainpower comes into existence, it will change everything, because it will just keep getting smarter and smarter. You are in essence saying, “Well if something with human-level intelligence comes along, nothing will change. We’ll still be ants, digging our burrows, feeding larva, foraging for food. The world will not change.” We may still be ants … but the world WILL change, whether we recognize it or not.
Its already happened, the AI is just staying quiet, just in case we noticed.( this then reminded me of Dr. Frankenstein, high in his castle, with his “AI”…and the townspeople outside, enraged, pitchforks at the ready:eek:)
Perhaps its the one giving everyone poor directions on their GPS…I’m going back to paper maps!!
We may have resource/ people bottlenecks that make the marvelous future slightly less marvelous, I’m thinking of the have/have nots issues. Plenty of scenarios where chaos becomes a bit of the norm, and scientific research takes a back seat to just staying alive.
Even if we don’t get human or super-human level AIs, other converging technologies would put into question chappachula (incredible IMHO) assertion that life in 2050 will be essentially the same as that in 2000. His prediction that he and Kurzweil will die ‘at age 85 or 90’ (and by extension so will the rest of us) is questionable, but even if it happens the difference will probably be the quality of life in the upper ages. People, after all, lived into their 90’s in the past, but the percentage was low and their quality of life was poor, even long before they got into their 90’s. In my folks parents generation, quality of life started dropping off quite a bit in their 60’s, and even in their 50’s. Today, we’ve pushed that back decades, and you see more an more people living to greater ages AND having more quality of life for longer as well. Who knows what advances in genetics, biology, medicine or even nano-tech will have on such predictions in the next few decades?
It’s laughable, to me, to assert things like ‘We’ll be using the same toilets , the same refrigerators , shopping at the same supermarkets, watching the same TV shows. We’ll still be going to work in an office, and working for a real, human boss, not some kind of mythical AI robot with superhuman brainpower .’, especially if one looks at the rather fundamental changes that happened in most of these things between 1950 and 2000 which, apparently, chappachula missed.
