In the future, as computers become more capable in terms of processing power, they will also be more capable in terms of simulations. Experiments and hypotheses are already simulated on today's computers, removing the need to physically test them. Things such as weather patterns and earthquakes are simulated on computers. So far, the majority of these simulations have been dealing with physics or other hard sciences.
Eventually, social sciences will also gain the advantage of simulation. Simulations regarding such subjects as sociology and political science will be ran on powerful computers. Given the massive amount of variables and relative complexity of behavioral sciences, they require much more processing power and information than simple (relatively speaking) natural and environmental simulations.
Imagine a computer program in which various societies, all with their own unique ideologies, are programmed to experience certain conditions. Political scientists can design governments and simulate them on various societies to find the strengths and weaknesses. If the society responds negatively to the governmental structure, it obviously needs revising. Conceivably, this could lead to much more perfect governments, as they have throughout history relied on trial and error (or brute force) to rule their people. Eventually, governments may be prepackaged after rigorous testing in simulations.
Sociologists could study these virtual societies and individuals, while applying various stimuli and monitoring the virtual being's reactions. To most accurately represent real humans, the environment in which the virtual humans "live" in would have to be as earth-like s possible, and the daily activities and ideologies would also require much attention in order to keep the simulations as accurate as possible.
Tuesday, February 22, 2011
Monday, January 3, 2011
Do AI's Need Us to Teach Them?
Artificial intelligence is still a future invention, yet the foundations for its existence may be present today. Nobody truly knows how AI will be created, but some have speculated that it may be developed through social networking sites. AI's will have their own profiles, and engage in social activities like friending and messaging. These primitive AI's will evolve cognitive behavior through interaction with humans on the sites.
Basically, social networking sites will serve as virtual environments where AI's can evolve. This speculation was done by Yury Millner, a Russian web tycoon quite familiar with the internet and how it works. So maybe it should be taken with a bit of respect.
We generally picture AI's as either intelligent or not, all or nothing. But we seldom consider that they will most likely be semi-intelligent for a period of time before they truly become intelligent. It is difficult to predict whether they will need to learn through interaction with humans (as in social networking), or if they will have the ability to self-teach. I believe that they will initially need to be taught through human interaction, but the first ones will be able to teach others. In a way, the first AI's will be the Adam and Eve of all future AI's. They will teach their predecessors how to learn and thus begin a virtual evolution.
This is similar to what I speculated a few months ago, so this only enforces my prediction that social networking sites will play a large role in the development of AI's and mind clones. The only question is when.
Basically, social networking sites will serve as virtual environments where AI's can evolve. This speculation was done by Yury Millner, a Russian web tycoon quite familiar with the internet and how it works. So maybe it should be taken with a bit of respect.
We generally picture AI's as either intelligent or not, all or nothing. But we seldom consider that they will most likely be semi-intelligent for a period of time before they truly become intelligent. It is difficult to predict whether they will need to learn through interaction with humans (as in social networking), or if they will have the ability to self-teach. I believe that they will initially need to be taught through human interaction, but the first ones will be able to teach others. In a way, the first AI's will be the Adam and Eve of all future AI's. They will teach their predecessors how to learn and thus begin a virtual evolution.
This is similar to what I speculated a few months ago, so this only enforces my prediction that social networking sites will play a large role in the development of AI's and mind clones. The only question is when.
Thursday, December 30, 2010
Will Simulating the Real World Create AI's?
With a new project announced, a team of international scientists have proposed building a simulation of the entire world. The Living Earth Simulator will not be the type of simulator one first pictures, as in a graphic virtual world. Rather it is a software program which operates at on supercomputers to comb through and analyze the large amounts of data on the internet for a selected output.
Through the use of semantic web technologies, the software will be able to more efficiently mine the vast archives of data on the internet. This effectively brings the internet to a more organized form and allows the massive amount of data to be used much more productively, rather than just staying stagnant while waiting for somebody to possibly pick them up in a search engine query, as they often do now.
While this simulator would undoubtedly be an effective data organizer and a huge advantage for scientists who need to simulate real world events for experiments, it could also be the first step towards creating environments in which AI can exist.
If software can "intelligently" comb the archives of the internet and produce given outputs, then the same software may be programmed to produce an AI. By searching the internet and finding all knowledge on computer science, mechanics, neural structure, and other relevant fields, the software could possibly have the capability to create the first true AI.
This project is sure to be years away and it would certainly be years after its initial availability that it would be sophisticated enough to create AI's, but this (or a later version of it) may be the missing link between us and the digital world that will ultimately create the long-coveted AI's.
Through the use of semantic web technologies, the software will be able to more efficiently mine the vast archives of data on the internet. This effectively brings the internet to a more organized form and allows the massive amount of data to be used much more productively, rather than just staying stagnant while waiting for somebody to possibly pick them up in a search engine query, as they often do now.
While this simulator would undoubtedly be an effective data organizer and a huge advantage for scientists who need to simulate real world events for experiments, it could also be the first step towards creating environments in which AI can exist.
If software can "intelligently" comb the archives of the internet and produce given outputs, then the same software may be programmed to produce an AI. By searching the internet and finding all knowledge on computer science, mechanics, neural structure, and other relevant fields, the software could possibly have the capability to create the first true AI.
This project is sure to be years away and it would certainly be years after its initial availability that it would be sophisticated enough to create AI's, but this (or a later version of it) may be the missing link between us and the digital world that will ultimately create the long-coveted AI's.
Wednesday, December 22, 2010
Savant World
We have all heard stories about people who can remember every detail of their life, and some who can even perform superhuman feats such as learn a language in a week, solve extremely complex mathematical problems, or recite passages from an array of literature, a la Dustin Hoffman in Rainman. These cases are however very far and few in between, making them extraordinary in their own right. With brain-enhancement advanced technologies, artificial consciousnesses (AC's) (and mind clones) to be developed within the next few decades, it may be possible that we will be living in a world of "savants."
I use the term savant, not necessarily in the traditional, biological nature, but rather for beings who can seamlessly harness the vast array of human knowledge through the internet and who can use technological cognitive enhancers to perform seemingly impossibly complex functions internally.
Take for example, there are people living today who have the natural ability to remember nearly every moment of their life. We call it photographic memory. It may not be too long before this ability will be common amongst people. I see two ways of this occurring.
The first way is through advanced lifelogging technology, as I've discussed before, in which every moment of one's life is digitally archived on computers. Along with the visual and audio archiving, thoughts and sensual experiences (however detailed) are also archived through brain-machine technology that monitors brain impulses. By syncing this information with the video and audio, this allows entire moments, not just images and movements, to be captured. Some method of digital recall will be implemented in the form of brain-machine devices connected to retinal-implants that allow one to switch from biological sight to "digital sight." The basic technologies for this recall are already being implemented. Microchips will presumably be able to replace natural light with digital images or create an augmented reality while a brain implant will trick the sense receptors into sensing the archived sensual information.
The other way for a photographic memory to become common in humans is through a more conventional means. That is through medical technology, such as cognitive enhancers and possibly neural-altering surgery. As medical technology advances, I expect cognitive enhancers to become increasingly common and may lead to an intelligence boom as more people expand their cognitive capabilities. I see the surgery prospect less plausible, as it would require a "miracle" breakthrough in the understanding of the brain to enhance cognitive function enough to the point that people will be willing to have brain surgery for it. It would also require a stunt in the growth of cognitive enhancers, which I see as unlikely.
Of course extreme cognitive enhancements will have an effect on the very nature of identity. Our definition and current understanding of identity may need to be rewritten once we achieve these abilities. Identity itself is built on the understanding of the world around us and our internal thoughts. Once we have a higher understanding of the world around us and have higher mental capabilities in which our thoughts are born, we will thus have dramatically altered the very nature of human identity. Identity is created through our cognitive ability. By my count, there are five elements of cognitive function, in which each one has its own impact on our identities. They are:
1) Technical processing (i.e. mathematics, engineering)
2) Abstract facts (i.e. historical dates, political events)
3) Abstract thought (i.e. ideologies, opinions)
4) Memory (i.e. personal events, world events)
5) Emotions (i.e. happiness, sadness)
Of these five elements, most of them are related. For example, emotions are generally influenced and thus defined by memories. It may be that once we have more control over our memories, we will thus have more control over our emotions. Our emotions may just be the mind's attempt to scrap pieces of memories together to make sense of the present, causing us to act and feel certain ways. Additionally, abstract thought is fundamentally the merger of memory, abstract facts, technical processing, and possibly emotions. These relations could go on, however the five elements are fundamentally different in their respective rights.
Transhumanism is often described as the point when humans become post-human, i.e. human 2.0. While I believe human 3.0 is when we become fully mechanical, implementing mind cloning technologies and AC's, I believe that 2.0 will only be achieved once all five cognitive elements are technologically enhanced. As I described earlier, memory will be dramatically altered this century with archiving technologies. Emotions will not be too difficult to alter, as we already have "mood drugs," and paired with memory enhancements, these will also be controllable.
Technical processing is another obstacle which poses not too big of a problem. Cognitive enhancers from medical technology will greatly improve technical processing, and eventually brain implants will ave the ability to speed up our brain's processing power.
Abstract facts are, in my opinion, going to be dealt with through seamless interaction with the internet. As we now can sit at a desktop, laptop, or smart phone and search the internet for facts, in a few decades we will have this ability wired directly to our brains. Possibly through brain-machine devices that read one's questions (through brain waves) and relay the answers either through retinal microchips or audio, facts will become easily attainable. Paired with memory enhancers, people will not forget or disregard these facts right after using them, as some argue people (particularly students) are doing now, thus making them dumber. Instead, people will be able to recall these facts even more seamlessly at later times from their memory.
Abstract thought, to me, is the last domino in the row. With enhancements in the other four elements, one will have more resources and processing power to create personal opinions, beliefs, and ideologies.
All of these technologies and prospects have been about enhancing our biological beings, yet I believe that we will possibly first see these functions being achieved by AC's, or mind clones. These beings will undoubtedly have the upper hand in terms of technical processing, abstracts facts, and memories soon after their development. It is abstract thought and emotions in which it may take a while longer before they can perform them as well as humans.
In the virtual environment in which the AC's will live, we may be able to control evolution towards achieving emotion and abstract thought rather quickly. In other words, we could favor the AC's that show signs of emotions and abstract thought, while destroying or not interacting with those that don't display these qualities. Thus, these qualities may be achievable if we understand how to correctly speed up artificial evolution.
Identity will undoubtedly be redefined, both for humans 2.0 and for AC's. Instead of thinking we know who we are and what we know, we will "know" who we are and what we know. In other words, our lives will be much more purpose-driven, detailed, and structured, rather than oblivious and unsure.
I use the term savant, not necessarily in the traditional, biological nature, but rather for beings who can seamlessly harness the vast array of human knowledge through the internet and who can use technological cognitive enhancers to perform seemingly impossibly complex functions internally.
Take for example, there are people living today who have the natural ability to remember nearly every moment of their life. We call it photographic memory. It may not be too long before this ability will be common amongst people. I see two ways of this occurring.
The first way is through advanced lifelogging technology, as I've discussed before, in which every moment of one's life is digitally archived on computers. Along with the visual and audio archiving, thoughts and sensual experiences (however detailed) are also archived through brain-machine technology that monitors brain impulses. By syncing this information with the video and audio, this allows entire moments, not just images and movements, to be captured. Some method of digital recall will be implemented in the form of brain-machine devices connected to retinal-implants that allow one to switch from biological sight to "digital sight." The basic technologies for this recall are already being implemented. Microchips will presumably be able to replace natural light with digital images or create an augmented reality while a brain implant will trick the sense receptors into sensing the archived sensual information.
The other way for a photographic memory to become common in humans is through a more conventional means. That is through medical technology, such as cognitive enhancers and possibly neural-altering surgery. As medical technology advances, I expect cognitive enhancers to become increasingly common and may lead to an intelligence boom as more people expand their cognitive capabilities. I see the surgery prospect less plausible, as it would require a "miracle" breakthrough in the understanding of the brain to enhance cognitive function enough to the point that people will be willing to have brain surgery for it. It would also require a stunt in the growth of cognitive enhancers, which I see as unlikely.
Of course extreme cognitive enhancements will have an effect on the very nature of identity. Our definition and current understanding of identity may need to be rewritten once we achieve these abilities. Identity itself is built on the understanding of the world around us and our internal thoughts. Once we have a higher understanding of the world around us and have higher mental capabilities in which our thoughts are born, we will thus have dramatically altered the very nature of human identity. Identity is created through our cognitive ability. By my count, there are five elements of cognitive function, in which each one has its own impact on our identities. They are:
1) Technical processing (i.e. mathematics, engineering)
2) Abstract facts (i.e. historical dates, political events)
3) Abstract thought (i.e. ideologies, opinions)
4) Memory (i.e. personal events, world events)
5) Emotions (i.e. happiness, sadness)
Of these five elements, most of them are related. For example, emotions are generally influenced and thus defined by memories. It may be that once we have more control over our memories, we will thus have more control over our emotions. Our emotions may just be the mind's attempt to scrap pieces of memories together to make sense of the present, causing us to act and feel certain ways. Additionally, abstract thought is fundamentally the merger of memory, abstract facts, technical processing, and possibly emotions. These relations could go on, however the five elements are fundamentally different in their respective rights.
Transhumanism is often described as the point when humans become post-human, i.e. human 2.0. While I believe human 3.0 is when we become fully mechanical, implementing mind cloning technologies and AC's, I believe that 2.0 will only be achieved once all five cognitive elements are technologically enhanced. As I described earlier, memory will be dramatically altered this century with archiving technologies. Emotions will not be too difficult to alter, as we already have "mood drugs," and paired with memory enhancements, these will also be controllable.
Technical processing is another obstacle which poses not too big of a problem. Cognitive enhancers from medical technology will greatly improve technical processing, and eventually brain implants will ave the ability to speed up our brain's processing power.
Abstract facts are, in my opinion, going to be dealt with through seamless interaction with the internet. As we now can sit at a desktop, laptop, or smart phone and search the internet for facts, in a few decades we will have this ability wired directly to our brains. Possibly through brain-machine devices that read one's questions (through brain waves) and relay the answers either through retinal microchips or audio, facts will become easily attainable. Paired with memory enhancers, people will not forget or disregard these facts right after using them, as some argue people (particularly students) are doing now, thus making them dumber. Instead, people will be able to recall these facts even more seamlessly at later times from their memory.
Abstract thought, to me, is the last domino in the row. With enhancements in the other four elements, one will have more resources and processing power to create personal opinions, beliefs, and ideologies.
All of these technologies and prospects have been about enhancing our biological beings, yet I believe that we will possibly first see these functions being achieved by AC's, or mind clones. These beings will undoubtedly have the upper hand in terms of technical processing, abstracts facts, and memories soon after their development. It is abstract thought and emotions in which it may take a while longer before they can perform them as well as humans.
In the virtual environment in which the AC's will live, we may be able to control evolution towards achieving emotion and abstract thought rather quickly. In other words, we could favor the AC's that show signs of emotions and abstract thought, while destroying or not interacting with those that don't display these qualities. Thus, these qualities may be achievable if we understand how to correctly speed up artificial evolution.
Identity will undoubtedly be redefined, both for humans 2.0 and for AC's. Instead of thinking we know who we are and what we know, we will "know" who we are and what we know. In other words, our lives will be much more purpose-driven, detailed, and structured, rather than oblivious and unsure.
Saturday, December 18, 2010
Programming for Creativity
The human mind is composed of numerous elements, all serving their respective purpose. It is understood that the right side of the brain is generally active in decision-making and the left side processes language. These two sides function cooperatively to produce what we consider to be a "working mind." However, the two sides may also be in a constant struggle for dominance, thus neither is able to reach its full potential.
In the case of mind clones/artificial consciousnesses (AC's), it should be possible to alter the ratio in which the two sides function. For example, one could have its "left side" processes turned down in order to give the right side processes more room to function. Once we have a more thorough understanding of the effects that each process has on others, we may be able to program the AC's specifically for certain tasks (i.e. art, mathematics, politics, etc.).
A recent report suggests that people who have damage to the left side of their brain are generally more creative than those with no damage. Regardless of the validity of this claim, I find it inarguable that shutting off certain parts of the brain, or certain functions, will alter the way in which the mind functions.
Thus, as we discover, through human tests or simulated experiments, what each brain section does and the effects they have on other parts of the brain, we will undoubtedly have the ability to customize our AC's to be the ultimate "whatever we want them to be."
The obstacle in which we must somehow overcome is the effects that shutting off one side of the "brain" in AC's will have on their virtual experience as a whole. It seems impossible that we could turn one side of a human brain down and then turn it back up without that human feeling as though something was wrong with his/her experience during that time.
If AC's are truly conscious, they may have the same incomplete feeling that humans would experience were one side to be turned downed. Possibly the solution would be to create AC's who do not function as true AC's, or only do until they have learned enough, then alter their process ratio for the remainder of their existence, effectively grooming them for their specific purpose.
In the case of mind clones/artificial consciousnesses (AC's), it should be possible to alter the ratio in which the two sides function. For example, one could have its "left side" processes turned down in order to give the right side processes more room to function. Once we have a more thorough understanding of the effects that each process has on others, we may be able to program the AC's specifically for certain tasks (i.e. art, mathematics, politics, etc.).
A recent report suggests that people who have damage to the left side of their brain are generally more creative than those with no damage. Regardless of the validity of this claim, I find it inarguable that shutting off certain parts of the brain, or certain functions, will alter the way in which the mind functions.
Thus, as we discover, through human tests or simulated experiments, what each brain section does and the effects they have on other parts of the brain, we will undoubtedly have the ability to customize our AC's to be the ultimate "whatever we want them to be."
The obstacle in which we must somehow overcome is the effects that shutting off one side of the "brain" in AC's will have on their virtual experience as a whole. It seems impossible that we could turn one side of a human brain down and then turn it back up without that human feeling as though something was wrong with his/her experience during that time.
If AC's are truly conscious, they may have the same incomplete feeling that humans would experience were one side to be turned downed. Possibly the solution would be to create AC's who do not function as true AC's, or only do until they have learned enough, then alter their process ratio for the remainder of their existence, effectively grooming them for their specific purpose.
Friday, December 17, 2010
Dialectical Process of Government
Humans have been running an experiment for around three millennium called government. Through this experiment, we have come up with a few systems, as described by Plato and other philosophers. These systems are monarchies, democracies, republics, anarchism, and others. Our experiment is still in progress, due to the fact that we have not yet created a government which truly meets the needs of its citizens.
This idea of a Utopia, in which everybody is happy and the government truly acts to fulfill the needs of all its citizens, has permeated philosophical novels for centuries. This illusive society, however common in literature, has never been able to exist in practice; at least not in large societies.
The reason for the lack of Utopian societies is undoubtedly due to the complexity and diversity of society, along with economic-political and human behavioral problems. From greed to selfishness, prejudice, fear, incompetency, and other unfortunate human traits, all have been at the root of government systems due to their human origin.
As societies grow larger, more complex, and more diverse, it is becoming increasingly harder for governments to adhere to the needs of all of their citizens. This is less a lack of resources as it is a fault in humans and our ability to work as a group towards meeting everybody's needs. This leaves us with two choices, either continue down the path of an increasingly widening gap between the rich and the poor, or redefine government and create a system in which everybody is taken care of. The problem may be that we are incapable of creating such a system.
As AI's and mind clones evolve, they will acquire the ability of independent analyzing and problem solving. As their computational power increases, so too does their analyzing ability. Thus, I believe we will eventually benefit by using them to create new government systems customized to the current times. In fact, in a century or so, we will have machine judges that may or may not work beside human judges, providing more accurate and unbiased decisions. We will also have machine politicians, working in ways similar to current politicians, but without the backstabbing and unproductive nature of current politicians. The machines will most likely work better together than humans and create more rational decisions than we do.
This may sound scary, but imagine a government in which decisions are based upon the advice given by an artificial intelligence. This AI will have all the information needed about the country to create rules and laws that provide the best living situations for the most people. Economies will be ran by AI's, along with foreign relations policies and other government aspects.
I, for one, believe that this will benefit humankind as a whole, as long as AIs stay unbiased and are not given full control of society. By this I mean that they will create laws and government systems, but humans should still maintain some sort of control to prevent the AI's from effectively taking over the world.
This idea of a Utopia, in which everybody is happy and the government truly acts to fulfill the needs of all its citizens, has permeated philosophical novels for centuries. This illusive society, however common in literature, has never been able to exist in practice; at least not in large societies.
The reason for the lack of Utopian societies is undoubtedly due to the complexity and diversity of society, along with economic-political and human behavioral problems. From greed to selfishness, prejudice, fear, incompetency, and other unfortunate human traits, all have been at the root of government systems due to their human origin.
As societies grow larger, more complex, and more diverse, it is becoming increasingly harder for governments to adhere to the needs of all of their citizens. This is less a lack of resources as it is a fault in humans and our ability to work as a group towards meeting everybody's needs. This leaves us with two choices, either continue down the path of an increasingly widening gap between the rich and the poor, or redefine government and create a system in which everybody is taken care of. The problem may be that we are incapable of creating such a system.
As AI's and mind clones evolve, they will acquire the ability of independent analyzing and problem solving. As their computational power increases, so too does their analyzing ability. Thus, I believe we will eventually benefit by using them to create new government systems customized to the current times. In fact, in a century or so, we will have machine judges that may or may not work beside human judges, providing more accurate and unbiased decisions. We will also have machine politicians, working in ways similar to current politicians, but without the backstabbing and unproductive nature of current politicians. The machines will most likely work better together than humans and create more rational decisions than we do.
This may sound scary, but imagine a government in which decisions are based upon the advice given by an artificial intelligence. This AI will have all the information needed about the country to create rules and laws that provide the best living situations for the most people. Economies will be ran by AI's, along with foreign relations policies and other government aspects.
I, for one, believe that this will benefit humankind as a whole, as long as AIs stay unbiased and are not given full control of society. By this I mean that they will create laws and government systems, but humans should still maintain some sort of control to prevent the AI's from effectively taking over the world.
Thursday, December 16, 2010
The First Stages of Mind Uploading
Often I speculate how exactly mind uploading will evolve and what path it will take before it becomes a reality. As I discussed in a previous post, one possibility is a "super Facebook," or a website that stores virtually all of one's personal qualities and quirks to the point of recreating the person's personality, or consciousness. This would require incredibly complex algorithms that generate personalities based on people's individual preferences and rely on sharing experiences, thus it may be impractical, if not impossible.
Another, perhaps more practical route would be in the form of brain-machine technology. Eventually, it will become common to connect our brains with computers and the vast resources of the internet. The majority of us will have visual and auditory recording devices embedded into our sensory organs, thus digitizing our senses. These devices will effectively record everything we see and hear (basically the mature version of lifecasting).
Additionally, we will be able to connect our brains to devices that monitor our reactions to external events. Through analyzing our responses to daily situations and having access to our memory banks in the form of recorded media, we will have made the first steps towards truly replicating human personalities.
It may be more difficult to create independently intelligent artificial beings, but these archived personalities will likely be our first steps toward creating truly virtual humans. Thus this is the path I believe mind uploading will take as it matures into a realized technology.
Another, perhaps more practical route would be in the form of brain-machine technology. Eventually, it will become common to connect our brains with computers and the vast resources of the internet. The majority of us will have visual and auditory recording devices embedded into our sensory organs, thus digitizing our senses. These devices will effectively record everything we see and hear (basically the mature version of lifecasting).
Additionally, we will be able to connect our brains to devices that monitor our reactions to external events. Through analyzing our responses to daily situations and having access to our memory banks in the form of recorded media, we will have made the first steps towards truly replicating human personalities.
It may be more difficult to create independently intelligent artificial beings, but these archived personalities will likely be our first steps toward creating truly virtual humans. Thus this is the path I believe mind uploading will take as it matures into a realized technology.
Subscribe to:
Posts (Atom)