Kurzweil Predicts Human Immortality
At the recent SCO6 supercomputing conference, Ray Kurzweil predicted human immortality by way of nanomedicine, and more. "Nanobots" in our bloodstreams will repair cellular and tissue damage as fast as it occurs. Affordable computers will be a thousand times more powerful than a human brain . . . Kurzweil clearly believes in "singularity now!"
Kurzweil as author, is an example of an uber-synthesist. He sifts through mountains of information to find technology trends. He predicts the future, without bothering to have tongue in cheek.
His biological predictions may even be timid, but Kurzweil's predictions of nanotech healers and machine super-cognition may be more than a little optimistic--from his point of view.
Humans are certainly not ready for super-intelligence, and computers have no point of reference. Human intelligence is based on emotions, which are based on biological survival drives. Computers have nothing similar to build on, other than (anthropomorphizing alert!) an unconscious "need" for electricity, more memory, and faster processors. As soon as computers begin to understand the need for speed and power, we may all be in trouble. Computers are not bound by billions of years of kludgy evolution--they can evolve exponentially.
So it comes down to: how soon can engineers/scientists build emotions and drives into computers--computers with actuators and power to act in the physical world? Because that is when Kurzweilian things start happening. Only computers can program massively parallel computers. They simply need a good reason to do so.
Kurzweil uncharacteristically provides a caveat: terrorists may learn to use this advanced technology. But more realistically, modern terrorists would like nothing more than to take human societies back to the stone age, where merely talking about Kurzweil's predictions would get your head chopped off.
Cross-posted at Al Fin Longevity.
According to Kurzweil, here's what we can expect in the not-so-distant future:Source.
—Doctors will be doing a backup of our memories by the late 2030s;
—By the late 2020s, doctors will be sending intelligent bots, or nanobots, into our bloodstreams to keep us healthy, and into our brains to keep us young;
—In 15 years, human longevity will be greatly extended. By the 2020s, we'll be adding a year of longevity or more for every year that passes;
—In the same timeframe, we'll routinely be in virtual reality environments. Instead of making a cell call, we could "meet" someone in a virtual world and take a walk on a virtual beach and chat. Business meetings and conference calls will be held in calming or inspiring virtual locations;
—When you're walking down the street and see someone you've met before, background information about that person will pop up on your glasses or in the periphery of your vision;
—Instead of spending hours in front of a desktop machine, computers will be more ingrained in our environment. For instance, computer monitors could be replaced by projections onto our retinas or on a virtual screen hovering in the air;
—Scientists will be able to rejuvenate all of someone's body tissues and organs by transforming their skin cells into youthful versions of other cell types;
—Need a little boost? Kurzweil says scientists will be able to regrow our own cells, tissues, and even whole organs, and then introduce them into our bodies, all without surgery. As part of what he calls the "emerging field of rejuvenation medicine," new tissue and organs will be built out of cells that have been made younger;
—Got heart trouble? No problem, says Kurzweil. "We'll be able to create new heart cells from your skin cells and introduce them into your system through the bloodstream. Over time, your heart cells get replaced with these new cells, and the result is a rejuvenated, young heart with your own DNA";
Kurzweil as author, is an example of an uber-synthesist. He sifts through mountains of information to find technology trends. He predicts the future, without bothering to have tongue in cheek.
His biological predictions may even be timid, but Kurzweil's predictions of nanotech healers and machine super-cognition may be more than a little optimistic--from his point of view.
Humans are certainly not ready for super-intelligence, and computers have no point of reference. Human intelligence is based on emotions, which are based on biological survival drives. Computers have nothing similar to build on, other than (anthropomorphizing alert!) an unconscious "need" for electricity, more memory, and faster processors. As soon as computers begin to understand the need for speed and power, we may all be in trouble. Computers are not bound by billions of years of kludgy evolution--they can evolve exponentially.
So it comes down to: how soon can engineers/scientists build emotions and drives into computers--computers with actuators and power to act in the physical world? Because that is when Kurzweilian things start happening. Only computers can program massively parallel computers. They simply need a good reason to do so.
Kurzweil uncharacteristically provides a caveat: terrorists may learn to use this advanced technology. But more realistically, modern terrorists would like nothing more than to take human societies back to the stone age, where merely talking about Kurzweil's predictions would get your head chopped off.
Cross-posted at Al Fin Longevity.
Labels: authors, biosingularity, immortality, life extension, Ray Kurzweil, Singularity
12 Comments:
Al Fin,
Rather than building emotions into computers, isn't it more likely that scientists and computer engineers will learn how to do download from computer to brain and vice versa? Isn't it likely that these download wil continue to become more and more comprehensive, until human beings are exactly what they are now, but thinking billions of times faster?
That's how I envision the future. Emotional thinking at the speed of light.
Tell me why I am wrong, please.
By the way, I very well may be proven wrong, but I honestly don't think computers, on their own, could ever truly exceed human intelligence. Why? Because, human intelligence includes will/creativity. Whatever we dream up, we eventually learn how to achieve. We are limited by time and space, but that just means it takes us time, and we have to work with matter.
There is nothing else like human intelligence this side of eternity.
Thing is, I can make these kinds of pronouncements and I won't be proven wrong until long after I'm dead. And, of course, by then, no one will know who I am.
:)
I have the confidence of historical anonymity.
One interesting thing about Kurzweil is the extent to which his life is dominated by his desire to live forever. His entire life is a regimen directed towards keeping his body alive long enough to enter the singularity. I think this may have begun to effect his analysis of the future. It may cause him to be too optimistic.
What do you think?
Those ideas are interesting, but when you get into the specifics of how minds and computers interface, uploading/downloading becomes very problematic. Likewise, emotions work through hormonal cascades in the brain and body. These events take time and have lingering effects.
Putting emotions and drives into computers is comparatively easy, a little at a time. Emotions will speed some computer capabilities and slow others, since an emotional computer will have to integrate emotions with other speedier functions.
Kurzweil wants to live forever and he has good reason. Only a few humans in history have been as prolific. He must enjoy both his inner and outer lives very much. When a human is healthy emotionally, intellectually, socially, and physically, he has no reason to want to die.
I'm not recommending wanting to die. Don't get me wrong.
Aren't hormones/emotions just chemicals that alter the way the neurons fire, or something like that?
Therefore, isn't the main thing reading the neurons, or some basic language the brain speaks, or are you sure the brain speaks many languages and hormone language is one of them?
Yes, hormones alter neuron firing and also alter glial (neuron support) function. The extrasynaptic "slow spreading" activity across regions of the brain is also influenced by neurohormones.
Neurons function in clusters, as synchronous waveforms that communicate with other synchronous waveforms, sometimes at different frequencies. The blog Develintel has good explanations for that phenomena.
There has been some early progress using neurochips to mimic some brain activity to facilitate brain regenerative activity. But the sheer complexity of neural net interactivity is mind-boggling!
I don't understand the processes well enough to understand what words to use to formulate my question, but I'll have a hack at it anyway?
Do all neurons across the brain (no matter what their function) fire within a certain limited frequency, and using similar chemical triggers?
Do they use a common language of sorts?
The reason I ask is, if they do, then isn't computer/brain communication a matter of interactin with neurons and not with hormones?
The brain's neurons are organized into interacting neural nets. The neuronal membership of these nets can change for different functions in the sense that sandlot teams can change from day to day by "choosing up sides." You might want to do a websearch for "neural synchrony." Here is one interesting look. Different parts of the brain apparently communicate through specific frequency oscillations--an intricate rhythmic dance honed by evolution over billions of years for survival purposes.
Brains are capricious, computers are deterministic. Vending machines function a lot like computers, if well stocked and maintained. If a vending machine worked like the brain, you would never know how much money to put in to get your soda--or whether you would get beer or piss dispensed instead of your soda.
Hormones modulate neural activity and can completely change the priorities of the conscious and sub-conscious. Our "will" is subject not only to our neuronal activity but to our hormonal activity. Our bodies contribute quite a lot to the state of our minds as well. Many people think more with their bodies than their brains.
Whatever type of machine is eventually developed that can feel pleasure and pain, can daydream, can feel the urgency of desire or hunger, and can be driven to kill to protect itself or things it loves--that machine will have little in common with any computers we can currently imagine.
For an interesting introduction to neural nets, look up Manfred Spitzer's "The Mind Within the Net."
So, let me see if I have this straight. A neural net is a grouping of neurons. The grouping itself is somewhat random. In other words, one could see the same image twice within an hours time and on each occasion a completely different set of neurons would be used in perceiving the image.
Is that right?
Hormones (or body thinking) cause the neural net formation to be even more random than if the perception were totally created by the brain.
Is that right?
So, the problem with using chip/brain interface is that perception is not done in any specific part of the brain. Therefore, a real chip/brain interface would have to have access to the entirety of the brain, and would have to be working at a tremendously high speed to keep up with the randomly forming neural nets.
Is that right?
A real brain/machine interface would certainly need access to the entire cortex, limbic system, hippocampus, thalamus, and sensory inputs, physiologic inputs, and autonomic activity.
But realistically, even if an identical brain to yours could be grown, if it lacked the "experiences" laid down over the years that your own brain possesses, there is no method known for "transferring" your "self" into the new brain--even if were anatomically identical. How much harder would it be to transfer your "self" into a machine that is completely alien in nature to your brain?
Brain function seems random in comparison to deterministic computers. But memories and thoughts in brain are distributed, unlike program data in computers which is stored at specific addresses.
Perception of an image is different from recall of that same image. But different recalls of the same image could certainly recruit somewhat different groupings of neurons. That is a good thing because it allows for "graceful" degradation of memory, rather than a system crashing loss of memory that occurs in computer memory degradation.
When machines are developed that can allow a similar "distributed" form of memory with graceful degradation, brain-like memories can be better stored by machine.
I don't believe hormone release increases randomness so much but instead "switches" the brain to different "states". Taking the image recall example, different hormone induced brain states might place the memory recall in different emotional connotations. One hormonal state might lead the image recall to bring a state of contentment, another hormonal state might place the person in a homicidal rage when recalling the same image.
As you can imagine, different parts of the brain would be called up for different emotional states, although the image being recalled is the "same".
It is not randomness so much as "complexity." Randomness is unpredictable--complex reactions can often be predicted, like the weather.
Emotions can be emulated in computers, and certain robot baby dolls simulate babies that need feeding, holding, comforting, etc. As time goes on, such emulation will become more complex. That type of programming in computers will become very convincing eventually, when computer architectures and algorithms grow enough in complexity.
Well, Al Fin. I thank you for explaining this to me. With my limited knowledge, I still have a hard time understanding why it is you believe that we won't eventually be able to create full interface with the human brain.
I understand your distinction between randomness and complexity. However, it seems to me that if complexity can be rather accurately predicted, then a computer would be able to interface with a certain high percentage of neurons involved in the net of a particular perception. Therefore, while the interface would not be perfect, it would still be capable of being captured and effective to a high degree.
Once again, I don't understand the science. I'm just thinking in image models.
Kurzweil seems confident that we will be able to interface between brain and computer to a very high degree within the next hundred years.
I am not disputing the eventuality of a functioning human-computer interfaces. I am only saying that putting emotions and emulating human-like thinking by "computers" will be easier, so will happen sooner.
Uploading and downloading entire human personalities is unlikely to happen in the next hundred years, in my opinion. But interfaces to computer "augments" will probably happen in the next twenty years.
Kurzweil is brilliant and has a lot of faith in the process of networked human invention, combined with increasing machine intelligence. Everything he predicts may come true. But probably not within his given time span.
Like you said, sometimes what a person wants to happen can affect what he predicts is going to happen.
Post a Comment
“During times of universal deceit, telling the truth becomes a revolutionary act” _George Orwell
<< Home