U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > General Forums > Science and Technology
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 03-01-2012, 08:29 PM
 
4,533 posts, read 6,924,233 times
Reputation: 2307

Advertisements

Quote:
Originally Posted by Woof View Post
Well sure, Night, but things have to start somewhere - you don't get a full scale robot to begin with, nor an omniscient personal assistant - and those beginnings have turned out to be quite popular and affordable. Roombas less than $200, iPhone 4s starting at $199 plus wireless and data plan.
LOL! Well, yeah, I agree that's true. Ya gotta start somewhere. But are you saying that Roomba was the start of domestic robots leading to the Technology Singularity as predicted? I think we could go much farther back regarding the start of robotics. Roomba is indeed an interesting gadget, but why bother paying $200 for one, not to mention the expectation of having to replace it every couple of years or so because they break down, assuming they last that long. You can bet there are plenty of fake Roombas being cranked out in China. And of course, there's Robomaid which looks like the one from the Jetsons. Depending on the size of the area to be cleaned, it can take quite a while between time spent actually sweeping and recharging the batteries to do something that could be done in 10-20 minutes or so with a broom and a vacuum cleaner. Roombas only have a small collection tray, and they're not going to remove and empty it by themselves. That's not counting the brush getting clogged from sweeping up hair, threads, etc.

The thing is that every major advancement in the history of mankind is just the most current version of many things that came before. Even the wheel is probably the offspring of previous items, probably logs to more easily move things around, but we can't point to a specific time when the first original idea came about. I would agree that technology is advancing more frequently, but I'd have a hard time claiming that a techological singularity will come about some time in the next 50 years. Some claim even sooner. In fact, I think describing it as a singularity is somewhat misleading.

I suspect there'll be some very impressive advancements over the next 50 or so years. Whether technology will surpass himan beings, described as a "Technological Singularity" with people trying to pinpoint to happen within that time frame is nothing more than wild speculation. We'll certainly see a continuation of knowledge and fast access to it. We might well see autonamous or integrated robotics as well. But I'd be hesitant to claim things will turn out as a singularity then. There are simply too many complex matters that can always get in the way of projected views. For example there were futurists who predicted that we'd all have flying cars, and people would be vacationing in space stations, or hotels on the Moon or Mars before the year 2000.

So I put forth the question, will technology reach a singularity, which in effect boils down to machines becoming smarter than humans, within the next 50 years? I'm not that optimistic, although I do think technology will continue producing more information much faster. Just because a machine may be faster and more accurate, doesn't necessariiy mean they'll be smarter. I do think there is a need for autonomous machines, especially in space exploration, but also as tools right here on Earth. We're already close to that although I don't think they'll be replacing humans any time soon, especially if there's a likelihood of machines overwhelming humans. They'll only be as good as they're programmed for. The human brain is incredibly complex. If the precise details of how the human brain works, then there's a good chance it could be replicated for machines.

I'd also put forth the question that if machines became more intelligent than humans, what possible reason would they have to remain on this planet?


Quote:
The reviews of Siri indicate that it's amazingly good at carrying on simple conversations relating to the task at hand using ordinary language (ie "give me a list of pizza parlors within a mile" or "remind me to call Debbie at noon concerning the Smith account"). That involves extraordinarily complex algorithms, I certainly didn't expect such a milestone to be passed within my lifetime.
I have no argument with blazing reviews praising the wonders of Siri, but still, that doesn't put it in the same category as a conversation among humans. Complex technology? Sure. But asking for list of nearby pizza parlors or a reminder to call someone, uses GPS to compare your location with a database listing of pizza parlors, etc., is impressive, but doesn't strike me as overly dumbfounding. It just represents a tool that's easier to use because it doesn't require reading the information. It all gets down to building a better mousetrap. Hmm, come to think of it, I'm not sure there's been much in the way of advancements in mousetrap technology for quite a while.
Reply With Quote Quick reply to this message

 
Old 03-01-2012, 09:22 PM
 
Location: where you sip the tea of the breasts of the spinsters of Utica
8,313 posts, read 10,627,990 times
Reputation: 7988
Well, anything involving data processing (such as conversations) is only a matter of complexity. A simple one-celled organism moves away from, or toward, food or light. Human conversations are a helluva lot more complex, but it's still data processing that could ultimately be traced back to neurons in different states (representing numbers in a similar way to the 1s and 0s of computers), firing off along certain pathways. If you ask humans for a list of nearby pizza parlors instead of Siri, they have to do a similar kind of computing.

As Siri type applications increase in complexity, they are going to resemble human conversations more and more, while of course having fast access to tremendously more information than people have.

By the way, I didn't intend the Roomba to be some sort of prime example of artificial intelligence - I just meant that it's something like that in a very primitive way, and that people DO like that sort of robotic maid assistance. It's considerably more economical than a maid, lay out $150 or so one time every few years, and you do little more than emptying the dirt.

Siri, on the other hand, is seriously impressive. Check out the reviews some time, especially from literate people such as in The New York Times. It can even do a little casual conversation not pertaining to numbers. I expect a next generation one to do even more casual "friendly" type conversations, and also to function as a general languages translator useful to travelers or international businessmen.
Reply With Quote Quick reply to this message
 
Old 03-02-2012, 09:40 AM
 
7,840 posts, read 8,159,678 times
Reputation: 9807
On a related subject, I posted this a while back in the work forum. How will all this robotic and AI advancement effect humans and work? Will we eventually become a society 'post-work'? Maybe not this lifetime, but eventually I can see virtually all work as being automated. Even the programming, design and maintenance will be automated by adaptive AI. And what will the fallout to the economic and govt systems be?

FuturistSpeaker.com – A Study of Future Trends and Predictions by Futurist Thomas Frey Blog Archive 2 Billion Jobs to Disappear by 2030

On robots, you've probably seen these but in case not I'll post them. Robots are becoming borderline creepy. Will be an interesting transition as we integrate them eventually into our lives.


ACTROID-F in AIST Open Lab 2010 02 - YouTube


Geminoid DK Mechanical Test - YouTube


PETMAN - YouTube


Freaky AI robot - YouTube


And one more for fun:


2011 Dodge Charger | The Future of Driving | Commercial - YouTube
Reply With Quote Quick reply to this message
 
Old 03-02-2012, 10:13 AM
 
Location: where you sip the tea of the breasts of the spinsters of Utica
8,313 posts, read 10,627,990 times
Reputation: 7988
Thanks, 11thHour. That fourth one is already way more advanced than the ones I saw early last year on youtube. It would be interesting to find out how many of his philosophical musings weren't directly programmed!
Reply With Quote Quick reply to this message
 
Old 03-02-2012, 03:50 PM
 
Location: Pueblo - Colorado's Second City
11,739 posts, read 18,317,370 times
Reputation: 3858
This is another advancement in AI. I don't think this was released but it shows how far we have come.

Reply With Quote Quick reply to this message
 
Old 03-02-2012, 05:16 PM
 
13,140 posts, read 33,759,227 times
Reputation: 11987
Quote:
Originally Posted by 11thHour View Post
Now that's friggin awesome .
Reply With Quote Quick reply to this message
 
Old 03-03-2012, 12:21 AM
 
Location: where you sip the tea of the breasts of the spinsters of Utica
8,313 posts, read 10,627,990 times
Reputation: 7988
Quote:
Originally Posted by Josseppie View Post
This is another advancement in AI. I don't think this was released but it shows how far we have come.

Wow, another friggin awesome one!

He seems capable of a little more realistic speech and emotion than Philip.
Reply With Quote Quick reply to this message
 
Old 03-03-2012, 01:59 PM
 
4,533 posts, read 6,924,233 times
Reputation: 2307
Quote:
Originally Posted by Woof View Post
Well, anything involving data processing (such as conversations) is only a matter of complexity. A simple one-celled organism moves away from, or toward, food or light. Human conversations are a helluva lot more complex, but it's still data processing that could ultimately be traced back to neurons in different states (representing numbers in a similar way to the 1s and 0s of computers), firing off along certain pathways. If you ask humans for a list of nearby pizza parlors instead of Siri, they have to do a similar kind of computing.
I'd take it a step farther and say that everything (not just conversations) involves information. I agree with you that human conversation is indeed an incredibly complex process. Interesting that you mention neurons. There's a thread from some time back that got into the workings of the brain, in particular the remarkable activity of the neural network which ultimately generates (among other things) perceptions, thought and ideas.

If I may jump back to the topic of the OP, it's not so much that I think it's impossible for artificial intelligence to be developed that could equal or exceed that of human intelligence. It's why hasn't the singularity gone mainstream, which I assume to mean at the present time.

At the present time it's a subject being explored, which is still very much in it's infancy, as you implied earlier by saying that it has to start somewhere. The issue I have a problem with is in terms of attempts to try to establish a time frame as to when such a Technological Singularity would take place. People like Kurzwell try to establish charts and suggest that such an event would occur within 25 to 50 years from now. The idea is based on the exponential rate at which technology is advancing, in particular, computer technology. That makes as much sense as saying global population will exponentially rise into the trillions by the next century. Paul Ehrlich attempted to do that in the late 60s with what he called the "Population Bomb". The thing is there are way too many factors that can get in the way in terms of date setting. It was thought we'd be vacationing on 2001 type torus space stations, colonizing the moon and Mars by the year 2000. It's been what, 40 years since anyone last set foot on the lunar surface, and there no plan to do so any time in the near future. all I'm saying is there are loads of unexpected and unforseen things that can change plans. Interestingly, such date setting, as in the Technological Singularity, tend to be just far enough off in the future, that it just amounts to guessing. The fact is that no one really knows. As it is, I won't likely be around at that time to see if it happens or not.

Will there continue to be remarkable technological advances in the next 25 to 50 years? Absolutely! Will it mark the predicted Technological Singularity? That's where it gets iffy.

We currently have supercomputers that can calculate a billion billion numbers per second, which in itself is mind staggeringly fast. But there's more than just number crunchng involved. While supercomputers may provide much faster and more accurate than someone trying to figure it out in their head, they are still not anything that can replicate the way the human brain works, even if what results from the brain isn't all that efficient or accurate.

I would think to be able to reach that level it'll take more than supercomputers. It might require highly advanced quantum computers. At the present time, although there have been advances in the development of quantum computers, it's far from anything practical at the present time. It too is very much in it's infancy. Who knows how it'll turn out or when it could be developed to a point that would make current computer technology obsolete. It could be within the next 50 years or sooner, or it could take a couple of hundred years or more. The potential is worth exploring, but the practical uses for it are not yet available.

The videos of robotics are very impressive, and so was the vid posted by Josseppie. Admittedly, in the area of robotics and AI, what is represented is still very primative and are not autonomous. The bots are still tethered to computers. There have been some bots (not humanoid) that have demonstrated sharing information remotely, which is greatly encouraging. One poster suggested that human-like android bots seem creepy. They do because they're still in the "uncanny valley" which is sort of a mix between human and cartoonish, both in physical features and movements and voice, which in effect means we have less empathy in terms of relating to them.
Uncanny valley - Wikipedia, the free encyclopedia


Okay, on with some cool stuff regarding AI. I think I posted these somewhere before, but they're worth posting here again.


Qbo Robot learns detects & recognizes - YouTube



Qbo meets Qbo - YouTube


The volume is pretty crappy in this vid, but it's worth seeing anyway. It involves learning remotely using cloud computing.

Qbo's learning and recognition in the cloud. Qbo robot - YouTube






Quote:
Siri, on the other hand, is seriously impressive. Check out the reviews some time, especially from literate people such as in The New York Times. It can even do a little casual conversation not pertaining to numbers. I expect a next generation one to do even more casual "friendly" type conversations, and also to function as a general languages translator useful to travelers or international businessmen.
That gets a big thumbs up from me. I've seen a few electronic hand-held audible language translators around, but some are not so great with some languages which make something like that a very daunting task. I speak both Thai and English which can be very complex, except for very basic phrases. But it would be excellent if such translators can be greatly improved for general communication (not that I have a need for one).
Reply With Quote Quick reply to this message
 
Old 03-03-2012, 04:44 PM
 
Location: Pueblo - Colorado's Second City
11,739 posts, read 18,317,370 times
Reputation: 3858
Default The singularity is near

Quote:
Originally Posted by NightBazaar View Post

At the present time it's a subject being explored, which is still very much in it's infancy, as you implied earlier by saying that it has to start somewhere. The issue I have a problem with is in terms of attempts to try to establish a time frame as to when such a Technological Singularity would take place. People like Kurzwell try to establish charts and suggest that such an event would occur within 25 to 50 years from now. The idea is based on the exponential rate at which technology is advancing, in particular, computer technology. That makes as much sense as saying global population will exponentially rise into the trillions by the next century. Paul Ehrlich attempted to do that in the late 60s with what he called the "Population Bomb". The thing is there are way too many factors that can get in the way in terms of date setting. It was thought we'd be vacationing on 2001 type torus space stations, colonizing the moon and Mars by the year 2000. It's been what, 40 years since anyone last set foot on the lunar surface, and there no plan to do so any time in the near future. all I'm saying is there are loads of unexpected and unforseen things that can change plans. Interestingly, such date setting, as in the Technological Singularity, tend to be just far enough off in the future, that it just amounts to guessing. The fact is that no one really knows. As it is, I won't likely be around at that time to see if it happens or not.
Computers have advanced at a exponential rate since the first "modern" computer was built in 1890. That is a fact and easily proven. So the question is will the trend continue or not? I have not seen any evidence to prove that it will stop, at least anytime soon, so there is no reason to think that Ray Kurzweil's models are wrong. Plus he is not the only one who forecasts out how computers will be as the entire gaming industry has done that for years so the games they bring on the market will be out at the right time to take advantage of the latest computer technology of the day.

Look at this graph on how fast supercomputers have advanced.




The link: File:Supercomputing-rmax-graph.png - Wikipedia, the free encyclopedia

This graph goes back 60 years and is a logarithmic graph that hides how fast the supercomputers have advanced but it is in a exponential rate. So does anyone think after all this time its just going to stop? No there is a logical reason for it as you use today's computers to build tomorrows computers so they are twice as fast then when tomorrow comes they use those to computers to build the next generation of computers and they are twice as fast and so on.

Finally I agree that it will take more then just computing power but a understanding of the human brain to have fully integrated AI. That is where the Blue Brain Project fits in and once complete, about 2019 and 10 years before Ray Kurzweil thought it would be, will enable AI to really take off. To be honest I find it fascinating that Ray seems to be on the conservative side of this issue but I can understand how that happened. When he was forecasting out he realized that computers that could fully simulate the brain would be affordable, yes some are out now but not really affordable for even the Blue Brain Project, would not be out till 2019. However what he did not anticipate was they do not need a computer to fully simulate the brain to complete most of the project just a computer that can simulate regions of the brain and they have that and had one since 2005. They are doing a bottom up approach starting with the Brain Stem and when they get it done and computers are advanced enough to simulate the entire human brain they will put it all together and complete their work and that should be by 2019. Once its done people will be like this came out of nowhere when in reality it has not and has been talked about by certain people, us included, for years before the "mainstream" public knew about it.

BTW I hope this made sense as I am ADHD so my thoughts can be allover the board and I know its not as concise as what you guys post but I don't want to spend hours editing a post and I think for the most part it gets across my point. At least I hope it does.

Last edited by Josseppie; 03-03-2012 at 05:10 PM..
Reply With Quote Quick reply to this message
 
Old 03-04-2012, 02:40 PM
 
4,533 posts, read 6,924,233 times
Reputation: 2307
Quote:
Originally Posted by Josseppie View Post
Computers have advanced at a exponential rate since the first "modern" computer was built in 1890. That is a fact and easily proven. So the question is will the trend continue or not? I have not seen any evidence to prove that it will stop, at least anytime soon, so there is no reason to think that Ray Kurzweil's models are wrong. Plus he is not the only one who forecasts out how computers will be as the entire gaming industry has done that for years so the games they bring on the market will be out at the right time to take advantage of the latest computer technology of the day.

Look at this graph on how fast supercomputers have advanced.


The link: File:Supercomputing-rmax-graph.png - Wikipedia, the free encyclopedia


This graph goes back 60 years and is a logarithmic graph that hides how fast the supercomputers have advanced but it is in a exponential rate. So does anyone think after all this time its just going to stop? No there is a logical reason for it as you use today's computers to build tomorrows computers so they are twice as fast then when tomorrow comes they use those to computers to build the next generation of computers and they are twice as fast and so on.
First of all, let me say thanks for your thoughtful reply and views.

If you look back at what I said, you'll see that I wrote "People like Kurzweil..." So it's not just Ray that holds that view, although he certainly seems to be touting it more. And why not? It sells books, provides speaking engagements, and generates a nice income. Just ask Erich von Daniken or Paul Ehrlich. To be fair, Ehrlich did ultimately back down and recognize that he was incorrect about his fervent predictions.

I'm not saying that charts don't show a steep increase in computer technology, nor am I saying that computer technology won't make significant improvements. However, the same thing can be said about charts regarding global population growth over the last 6000-8000 years. In fact the population growth charts overall show a much steeper upward curve over the last few centuries than computer technology. Will global population continue to increase at exponential rates? No. Eventually, we'd run out of room and resources. And that's not counting wars, disease, natural catastrophies, starvation, and so on.

If we rely solely on charts and trends to predict what will happen in the future, then it really is nothing more than wishful thinking and speculation. The reason is because, like it or not, there are plenty of obstacles that can and do get in the way that can abruptly change things. That was why I also said that despite predictions that we'd have thriving colonies on the Moon, etc., before the year 2000, that clearly has gone into a much different direction than had been expected or anticipated. It doesn't mean we won't continue to advance, but it does mean that exponential growth trends, whether it's population growth or computer technology, cannot continue upward indefinitely. Sometimes trends level off and sometimes they can take a nosedive.

One problem we often have is putting too many eggs into only a few baskets. If the baskets should tear, the eggs are going to be a splattered mess.

If computer technology is to make any truly radical changes, we'll probably have to develop a very different technology to do it, which you seem to agree with. Quantum computing could provide much faster and more accurate computing, but even so, there are no guarantees it will provide us with solutions to all our problems. In any case, we're far from having quantum computers that work as efficiently as our current computers, which includes supercomputers.




Quote:
Finally I agree that it will take more then just computing power but a understanding of the human brain to have fully integrated AI. That is where the Blue Brain Project fits in and once complete, about 2019 and 10 years before Ray Kurzweil thought it would be, will enable AI to really take off. To be honest I find it fascinating that Ray seems to be on the conservative side of this issue but I can understand how that happened. When he was forecasting out he realized that computers that could fully simulate the brain would be affordable, yes some are out now but not really affordable for even the Blue Brain Project, would not be out till 2019. However what he did not anticipate was they do not need a computer to fully simulate the brain to complete most of the project just a computer that can simulate regions of the brain and they have that and had one since 2005. They are doing a bottom up approach starting with the Brain Stem and when they get it done and computers are advanced enough to simulate the entire human brain they will put it all together and complete their work and that should be by 2019. Once its done people will be like this came out of nowhere when in reality it has not and has been talked about by certain people, us included, for years before the "mainstream" public knew about it.
Whoa, slow down there partner. Kurzweil's predictions are centered around a "technological singularity", a point at which the world will radically change because of a shift of automonous intelligence in computers that will exceed that of human intelligence. You mentioned "Project Blue Brain" as if to validate the idea that a technological singularity is approaching sooner than Kurzweil expected. Forget about Kurzweil. Those on the front lines are saying all that's being done at the present time is an ambitious plan to map out the human brain.

At present, only the most primative part of the brain, the cortex, has been mapped. The plan to hoped to gradually map out other areas of the brain with the ultimate goal of having mapped the entire human brain to have a simulation which can be useful for several applications, including (but not limited to) medical applications. It'd be a lot easier to practice on a simulated brain than a real one. Robotics is another application. The other stages are intended to include what has been previously mapped but at increasing volumes anywhere from incorporating 100's of times that of previous findings to 1000's of times each step of the way. And that's just mapping. Here's a look at what goes on in the brain. Keep in mind that there are about 100 billion neurons and trillions of combinations. The way the the brain processes information is pretty remarkable that it ultimately results in thoughts, perceptions, and ideas that enable us to visualize abstracts and respond to world around us.



Neurons and how they work - YouTube



Now, let's see what those on the front lines of mapping the brain, like Henry Markham (the director of Project Blue Brain) and Idan Seveg of Israel have to say. How long will it take? No one knows for sure, but it could take a few decades or it could take a hundred or more years. Are there any obstacles that could slow things down? Yes. It all relies on funding of very large amounts of money, much of it coming from governments, but also grants and private funding. If the funding slows down or stops, for any of numerous reasons, the project slows down or stops. With the state of how the global economy is today, it's hard to guess. It gets down to economic priorities. There have been plenty of casualities, like the Tevatron collider, the space shuttle, human landings on the Moon, etc. It happens. Things don't always turn out as planned or hoped for.



The Blue Brain Project - YouTube



Henry Markram: Simulating the Brain





Quote:
BTW I hope this made sense as I am ADHD so my thoughts can be allover the board and I know its not as concise as what you guys post but I don't want to spend hours editing a post and I think for the most part it gets across my point. At least I hope it does.
No need to apologize for anything. You're not alone. We all tend to wander with our thoughts, and are not always so concise and accurate. You're doing just fine, although I think it helps to understand that (in this subject) it's extremely complex and there are no simple answers. Not only does it offer the potential of some greatly impressive benefits, but there are also numerous obstacles to take into consideration before declaring what will happen in the future and when it will happen.

If I may, allow me to pose a couple of interesting questions. Let's say that indeed artificial intelligence can be achieved that actually exceed that of humans. And let's say that artificial intelligence is completely autonomous. What we would then have is an advanced intelligence based on technology and mechanics rather than biological. A new form of life.

First of all, who do you think would be first in line to grab on to it? I wouldn't be surprised that it would probably fall into the hands of the military. Issac Asimov originally established the "Laws of Robotics", which is quite reasonable as a way to achieve and ensure peaceful uses. The idea is that autonomous robots would be designed to prevent them from harming humans. But in the world we live in today, I think it's pretty clear that there are people who have less than peaceful motives. What would happen if this technology falls into the wrong hands like extremists? You can bet yer boots that would eventually happen. As it is, militaries are already looking at the possibility of using autonomous robots as war machines. Would that be okay with you?

Secondly, if an autonomous AI system were much more highly advanced than any human, what would prevent that system from deciding that humans are too inefficient, and decide to either abandon the planet and let humans work out their own problems, or in a worst case scenario, to either force humans to integrate, or simply decide that humans are useless and that it'd be better to exterminate all of them? The idea here being that AI systems become far more advanced than humans. How would you feel about such a scenario?

I'm not saying that's how it will end up, but how could such a possible outcome be prevented? International treaties are meaningless to those who aren't willing to take part in abiding by them. I think it's healthy to look at all sides of the issue, to look at the negatives as well as the positives.
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply

Quick Reply
Message:


Over $99,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > General Forums > Science and Technology
Similar Threads

All times are GMT -6.

2005-2017, Advameg, Inc.

City-Data.com - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32 - Top