Is the Human Era Facing Extinction?

[paulpatter]

It's called Singularity and is defined as "The moment when technological change becomes so rapid and profound, it represents a rupture in the fabric of human history" (Time magazine, Feb. 21, 2011). Lev Grossman, writing in the Time article, states that computers are becoming faster faster, and that artificial intelligence will become a reality. His primary source is Raymond Kurzweil, who invented the computer than translates the printed word into speech. Grossman reports: ". . .there's no reason to think computers [will] stop getting more powerful. They [will] keep on developing until they [are] far more intelligent than we are." That means, according to another Singularity researcher, that "Within 30 years, we will have the means to create superhuman intelligence. Shortly after, the human era will be ended" (Vernor Vinge, The Coming Technological Singularity). Grossman emphasizes, based on his research, that Singularity isn't science fiction, "no more than a weather forecast is science fiction. . .it's a serious hypothesis about the future of life on earth."

It's a fascinating article, one that caused me to reflect on my son's patriarchal blessing--given 20 years ago--that (paraphrasing) "the world will experience an explosion of information, and you will live to see it."

In any event, I encourage you to read the article. I'm not saying it's the "gospel truth," but it's certainly worth a serious--if not sobering--look.

[nicolasconnault]

It's called Singularity and is defined as "The moment when technological change becomes so rapid and profound, it represents a rupture in the fabric of human history" (Time magazine, Feb. 21, 2011). Lev Grossman, writing in the Time article, states that computers are becoming faster faster, and that artificial intelligence will become a reality. His primary source is Raymond Kurzweil, who invented the computer than translates the printed word into speech. Grossman reports: ". . .there's no reason to think computers [will] stop getting more powerful. They [will] keep on developing until they [are] far more intelligent than we are." That means, according to another Singularity researcher, that "Within 30 years, we will have the means to create superhuman intelligence. Shortly after, the human era will be ended" (Vernor Vinge, The Coming Technological Singularity). Grossman emphasizes, based on his research, that Singularity isn't science fiction, "no more than a weather forecast is science fiction. . .it's a serious hypothesis about the future of life on earth."

It's a fascinating article, one that caused me to reflect on my son's patriarchal blessing--given 20 years ago--that (paraphrasing) "the world will experience an explosion of information, and you will live to see it."

In any event, I encourage you to read the article. I'm not saying it's the "gospel truth," but it's certainly worth a serious--if not sobering--look.

There is nothing intelligent about computers. Never was, and probably never will be. The reason is simple: we don't understand how human intelligence works. Until we understand that, how can we possibly design machines that reproduce or exceed human capacity? Sure, computers can process and store information far better and faster than humans, but that is not the definition of intelligence. Self-awareness is necessary, and that, for computers, is pure science-fiction.

[TAO]

Nah, self-awareness is possible... what will be difficult is writing a program which knows how to write programs which know how to program.

Right now we know how to write a program which writes other programs.

We need a step above that to get complex intelligent AI to work. A program which writes programs which can write other programs.

And we might need a step above that... or even more... so 1 or more steps than we have right now.

[USU78]

He sprang from the cabin window as he said this, upon the ice raft which lay close to the vessel. He was soon borne away by the waves and lost in darkness and distance.

[Calm]

It's so nice when I know what other people are talking about before I turn to google.

Besides self-awareness, creativity and several other hard to quantify are aspects of human intelligence. At this point it is easy to envision a computer that can calculate outcomes quickly enough that it appears to be creative (because it can test so many modeled options so quickly to come up with something that is of value)...but unless it can actually create new parameters for its own programming, the creativity (and other stuff) will only be a mimic and not the actual thing. Human intelligence has the ability to take a variety of apparently unconnected things and place them together effectively without having to test a million of other items in order to come up with ones that might work. While programs for pattern recognition, etc. can help with this aspect, it would have to take a fundamental change in abilities in the computer in order to truly emulate the human thought process as thought is not just a combination of yes no decisions, even if most thought probably can be reduced to that model.

And as long as irrationality is part of the human thought process, I am not sure how you could get a machine to effectively mimic that as pure randomness would not yield a similar result since irrationality is rarely random.

In the example given of the computer having written the music piece, the creativity was not the computers, but those who designed the program as the computer was just following their instructions, not creating his own ideas and testing them.

Haven't finished the article yet, so perhaps it addresses this issue. I find there is a huge assumption (and a very common one) on which the whole article is resting...and it is most obvious in this comment:

We will successfully reverse-engineer the human brain

The assumption is that human intelligence is solely a result of the brain machine (plus environmental influences). If, however, the brain serves a significant but only partial role and there are other aspects that are less amendable to be duplicated by computer functions, this could cause a slowdown or even an insurmountable obstacle to overcome.

More 'man is a machine' assumptions:

"It's just childish. The human body is a machine that has a bunch of functions, and it accumulates various types of damage as a side effect of the normal function of the machine. Therefore in principal that damage can be repaired periodically. This is why we have vintage cars. It's really just a matter of paying attention. The whole of medicine consists of messing about with what looks pretty inevitable until you figure out how to make it not inevitable."

I see some more realistic, in my view, positions later in the paper:

The entire field of artificial intelligence, or AI, is devoted to this question. But AI doesn't currently produce the kind of intelligence we associate with humans or even with talking computers in movies

[nicolasconnault]

And don't "forget" that the human brain has an unlimited storage capacity... And scientists have no idea how that is possible, since the size of a person's accumulated knowledge is not proportional to that person's brain size.

[kolipoki09]

We discussed this very article tonight during a meeting of the BYU-Idaho Philosophical Society. Our "token" math major whom I will refer to as Noumena, believed he could develop an algorithm capable of producing artificial intelligence that would be able to transcend human ethics. Either he reads too many math books for fun, or he just watched The Social Network and thought the word "algorithm" sounded cool so he started using it. At this point, both options are feasible.

Perhaps Facebook will somehow vastly inflate through technological nucleosynthesis, implode on itself and create a black hole. Or maybe someone will just take a leap of faith and google Google to save us all the trouble.

I'll take it a day at a time. And when the time comes around later in life when this prediction will either become moot or written in stone, I'll look forward to watching. Perhaps by then I'll have enough posts here to be a "Contributor."

[EDIT] 600 Posts in less than three years! WAHOO!!!

[BCSpace]

Man made machines make music for the man

Now machines make music while the man makes plans

A second generation, a generation lost

While the man prepares for the holocaust

Man made machines to control the days

Now machines control while the man obeys

A second generation, a generation lost

While machines prepare for the holocaust

[Analytics]

Besides self-awareness, creativity and several other hard to quantify are aspects of human intelligence. At this point it is easy to envision a computer that can calculate outcomes quickly enough that it appears to be creative (because it can test so many modeled options so quickly to come up with something that is of value)...but unless it can actually create new parameters for its own programming, the creativity (and other stuff) will only be a mimic and not the actual thing. Human intelligence has the ability to take a variety of apparently unconnected things and place them together effectively without having to test a million of other items in order to come up with ones that might work. While programs for pattern recognition, etc. can help with this aspect, it would have to take a fundamental change in abilities in the computer in order to truly emulate the human thought process as thought is not just a combination of yes no decisions, even if most thought probably can be reduced to that model.

And as long as irrationality is part of the human thought process, I am not sure how you could get a machine to effectively mimic that as pure randomness would not yield a similar result since irrationality is rarely random.

In the example given of the computer having written the music piece, the creativity was not the computers, but those who designed the program as the computer was just following their instructions, not creating his own ideas and testing them.

Haven't finished the article yet, so perhaps it addresses this issue. I find there is a huge assumption (and a very common one) on which the whole article is resting...and it is most obvious in this comment:The assumption is that human intelligence is solely a result of the brain machine (plus environmental influences). If, however, the brain serves a significant but only partial role and there are other aspects that are less amendable to be duplicated by computer functions, this could cause a slowdown or even an insurmountable obstacle to overcome.

More 'man is a machine' assumptions:

I see some more realistic, in my view, positions later in the paper:Even with this it is still man being treated as a biological machine...which may or may not be true.

I find the predictions on the extension of life more likely.

I would put my money on Kurzweil. A machine would probably never be conscious in a philosophical sense, but machines are able to observe, analyze, decide, and act. Sure, there might be "things going on in our brains that can't be duplicated electronically no matter how many MIPS you throw at them," but that's besides the point. The point is that the power of computers is growing exponentially, and nothing has been identified that will cause that to change. That trajectory doesn't lead to computers becoming like the human brain--it leads to it surpassing it at a breathtaking speed.

It's a fascinating article.

[Nofear]

There is actually an organization that looks to/for the singularity through a Mormon lens: Mormon Transhumanist Association.

I personally disagree with the theological interpretations of many of the participants, but do generally agree with the attitude that technology is a potential force for immense good and potentially transformative influence.

Stepping outside the LDS bubble, we see that there is much interest in the transhumanist concepts. Consider, for example the recent founding of Singularity University it founders and partners.

PS: I haven't actually read the article and think Time did a horrible job with their cover art.

PPS: Being a libertarian (a free will philosophy) I believe a true AI is fundamentally impossible.

[nicolasconnault]

I would put my money on Kurzweil. A machine would probably never be conscious in a philosophical sense, but machines are able to observe, analyze, decide, and act. Sure, there might be "things going on in our brains that can't be duplicated electronically no matter how many MIPS you throw at them," but that's besides the point. The point is that the power of computers is growing exponentially, and nothing has been identified that will cause that to change. That trajectory doesn't lead to computers becoming like the human brain--it leads to it surpassing it at a breathtaking speed.

It's a fascinating article.

Exponential growth in a particular direction doesn't make it magically cross a boundary of a different dimension (namely "intelligence"). The mistake made here is to break down human intelligence into its components (memory, processing speed, data retrieval efficiency etc.) and to observe how well computers do each of them. We marvel at how computers perform on some of these dimensions, and at the exponential growth at which they are mastering these components of intelligence, but they are still at square 1 in many other components of intelligence. For example, computers are incapable of acquiring information by evaluating testimonial data (judging whether or not an outside statement is believable or not, worth accepting or not). They completely lack any sort of intuition, which is an essential characteristic of human intelligence.

So computers may surpass the human brain in some areas, but are absolutely inept in others. This is old news. When the first electronic calculator was invented, this already applied.

[paulpatter]

There is nothing intelligent about computers. Never was, and probably never will be. The reason is simple: we don't understand how human intelligence works. Until we understand that, how can we possibly design machines that reproduce or exceed human capacity? Sure, computers can process and store information far better and faster than humans, but that is not the definition of intelligence. Self-awareness is necessary, and that, for computers, is pure science-fiction.

Some exceptionally knowledgeable people appear to disagree with you, including NASA, which hosts three-year old Singularity University (offers interdisciplinary courses for graduates and executives); as well as Google's Larry Page, SU's CEO and co-founder. Grossman writes: "There's an intellectual gag reflex that kicks in anytime you try to swallow an idea that involves superintelligent immortal cyborgs, but suppress it if you can, because while. . .Singularity appears to be, on the face of it, preposterous, it's an idea that rewards sober, careful evaluation."

The article points out that humans tend to think in a linear fashion, but technological progress happens exponentially: "Exponential curves start slowly, then rocket skyward toward infinity." According to Kurzweil, by 2045, "given the vast increases in computing power and the vast reductions in the cost of same, the quantity of artificial intelligence created will be about a billion times the sum of all the human intelligence that exists today." The article characterizes Kurzweil as saying that "there is no fundamental difference between flesh and silicon that would prevent the latter from thinking. He defies biologists to come up with a neurological mechanism that could not be modeled or at least matched in power and flexibility by software running on a computer."

Please note that I'm just the messenger here. Clearly, however, the message merits consideration--and criticism as well.

[Analytics]

Exponential growth in a particular direction doesn't make it magically cross a boundary of a different dimension (namely "intelligence"). The mistake made here is to break down human intelligence into its components (memory, processing speed, data retrieval efficiency etc.) and to observe how well computers do each of them. We marvel at how computers perform on some of these dimensions, and at the exponential growth at which they are mastering these components of intelligence, but they are still at square 1 in many other components of intelligence. For example, computers are incapable of acquiring information by evaluating testimonial data (judging whether or not an outside statement is believable or not, worth accepting or not). They completely lack any sort of intuition, which is an essential characteristic of human intelligence.

So computers may surpass the human brain in some areas, but are absolutely inept in others. This is old news. When the first electronic calculator was invented, this already applied.

I don't agree with your example of evaluating testimonial data

[nicolasconnault]

Some exceptionally knowledgeable people appear to disagree with you, including NASA, which hosts three-year old Singularity University (offers interdisciplinary courses for graduates and executives); as well as Google's Larry Page, SU's CEO and co-founder. Grossman writes: "There's an intellectual gag reflex that kicks in anytime you try to swallow an idea that involves superintelligent immortal cyborgs, but suppress it if you can, because while. . .Singularity appears to be, on the face of it, preposterous, it's an idea that rewards sober, careful evaluation."

The article points out that humans tend to think in a linear fashion, but technological progress happens exponentially: "Exponential curves start slowly, then rocket skyward toward infinity." According to Kurzweil, by 2045, "given the vast increases in computing power and the vast reductions in the cost of same, the quantity of artificial intelligence created will be about a billion times the sum of all the human intelligence that exists today." The article characterizes Kurzweil as saying that "there is no fundamental difference between flesh and silicon that would prevent the latter from thinking. He defies biologists to come up with a neurological mechanism that could not be modeled or at least matched in power and flexibility by software running on a computer."

Please note that I'm just the messenger here. Clearly, however, the message merits consideration--and criticism as well.

Interesting stuff indeed. However, do you know of any exponential growth that has continued exponentially ad infinitum? I am not convinced that processing speeds will continue to increase exponentially forever.

What I am concerned about, is that people like Kurzweil seem to place a much higher value on certain aspects of human intelligence, while neglecting others. His theories are rooted in the information processing model of human intelligence, which is reductionist in nature, just as all models are. "Artificial Intelligence" is such a popular term that it tends to redefine how we think about human intelligence, based on what we are achieving with AI. It is much easier to understand a system that we created ourselves (AI) than the human brain, which we did not create. Information processing became popular during the advent of the computer era, and has been used to try to make sense of human cognition from a computer design's perspective ever since, often very usefully so, sometimes tragically not.

[nicolasconnault]

I don't agree with your example of evaluating testimonial data

[Jason]

Lev Grossman, writing in the Time article, states that computers are becoming faster faster, and that artificial intelligence will become a reality.

An interesting statement, in that it appears to be faith-based.

His primary source is Raymond Kurzweil, who invented the computer than translates the printed word into speech. Grossman reports: ". . .there's no reason to think computers [will] stop getting more powerful. They [will] keep on developing until they [are] far more intelligent than we are." That means, according to another Singularity researcher, that "Within 30 years, we will have the means to create superhuman intelligence. Shortly after, the human era will be ended" (Vernor Vinge, The Coming Technological Singularity). Grossman emphasizes, based on his research, that Singularity isn't science fiction, "no more than a weather forecast is science fiction. . .it's a serious hypothesis about the future of life on earth."

Like Global Warming is a serious hypothesis?

[Analytics]

Interesting stuff indeed. However, do you know of any exponential growth that has continued exponentially ad infinitum? I am not convinced that processing speeds will continue to increase exponentially forever.

When something is driven by an exponential growth process, it always stops growing exponentially, but not until an internal or external factor stops it. What

[Jeff K.]

The Human race is doomed, we lost to a computer on Jeopardy.

[thesometimesaint]

There is always the off button.

Ps. http://www.imdb.com/video/screenplay/vi329318937/

[CV75]

the human era will be ended

I think this is clearly in progress as more people are tethered to their information and entertainment devices--some can hardly relate or attend on a human-to-human level anymore. It is also progressing as more people treat others badly with less and less impunity.

[Jeff K.]

Today its cell phones and blackberrys and androids. Tomorrow, the borg phone.

[nicolasconnault]

When something is driven by an exponential growth process, it always stops growing exponentially, but not until an internal or external factor stops it. What

[Mola Ram Suda Ram]

So a..... I have a burning question, that I know is on everyone's mind: "When is Skynet going to go on line?"

[nicolasconnault]

I think this is clearly in progress as more people are tethered to their information and entertainment devices--some can hardly relate or attend on a human-to-human level anymore. It is also progressing as more people treat others badly with less and less impunity.

I think you mean "more and more impunity"

[Analytics]

So a..... I have a burning question, that I know is on everyone's mind: "When is Skynet going to go on line?"

According to TIME magazine, it will be in the year 2045.