# Introducing the Future

Discussion in 'Intelligence & Machines' started by Cris, Feb 26, 2001.

Not open for further replies.
1. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
As computing power reaches that of the human brain and beyond we will be faced with the inevitability that machines will have intelligence greater than our own. Is this good or bad, how will AI machines view us? Will they see us in the same light as we currently view apes. Perhaps we think that we can control this technology. I don't see that that will be possible. The road to AI machines that can outthink us is inevitable, we must come to terms with that and deal with it.

Cris

Last edited: Feb 27, 2001

3. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
And 20 Years Out? - Another Perspective.

(Extract from The Rapidly Changing Face of Computing, jeff.harrow@compaq.com)

I'm pretty clear that I don't really know what things will be like twenty years from now -- all I have to do is think back to twenty years ago (a year before the first IBM PC) to realize that even my optimistic view of technology never envisioned my current 1 GHz PC with its third-of-a-gigabyte of memory and 40 gigabytes of disk space and broadband connection to a global communications network. So, especially considering that the rate of technology advancement is growing each year, how could I possibly guess at twenty years from now?

But that doesn't stop some people from trying their best to extrapolate where today's science, and tomorrow's ideas, will take us, and RCFoC reader Tony Waltham brings our attention to a series of future-looking articles from British Telecom that I found interesting, if occasionally disquieting (http://www.bt.co.uk/bttj/tomorrow/index.htm). I'm not sure that I want to see all of these articles' prognostications come true, but they certainly are valuable food for thought.

For example (with a bit of my interpretation thrown in), the first article by BT's Ian Pearson peers through his crystal ball to see computer intelligence catching up with human intelligence by 2020 -- essentially, synthetic intelligence will be able to pass the Turing Test (you won't be able to detect if you're conversing with a human or an Artificial Intelligence -- see Ray Kurzweil's recent books for more on this subject.) Such AIs will be so good that they will far surpass the 1980s Eliza "toy" psychoanalyst software, and will actually pose as effective robotic psychiatrists.

Virtual environments, such as in today's game EverQuest (www.everquest.com), will become fully immersive 3D environments for both play and work, bringing (dare I say it) a new dimension to telecommuting and virtual meetings. And they will also bring new business opportunities. For example, if we're going to be represented by our avatars in virtual environments, won't we want them to look their best? The purchase or rental of virtual fashions may spark a very real industry (in fact it's already done so in early virtual communities, such as in EverQuest's Norrath, where people work long and hard to get just the right colored robe, or that perfect flashy armor, that they want to be seen wearing.)

Sensors, such as the growing number of municipal cameras that watch vehicular and pedestrian traffic, will be augmented by a vast number of tiny "smart dust" sensors that will diligently fill up the greatly enhanced storage devices of 2020. Everything will be connected, anywhere. Indeed, Ian sees 75% of the global population being able to reach out and touch the Internet. (Of course, he recognizes that not everyone will be pleased with this level of technological intrusion; he also envisions Luddite communities growing in protest.)

By 2020, he sees the global Internet spelling the doom of national currencies. Much as the Euro is becoming the common coin of the European realm, Ian sees a global electronic currency similarly becoming the coin of the virtual realm. And then, because of the Internet's reach into the real world, he sees it becoming the global currency of choice. (Think that's improbable? E-gold (www.e-gold.com) is working right now to establish just such specie!)

Timeline.

And there's more to whet our technological whistle. BT offers us a year-by-year timeline of innovations over the next 20 years. They include (in small part), some things that have already happened, such as Casio's wristwatch camera and speech dialing. But they go on to suggest things like "single sheet PCs with processing built into the display" by 2002, synthetic organic life by 2003, 300 gigabit CD-ROMs by 2004, and on-line voting in the UK by 2007 (which might have saved us quite a bit of angst on this side of the pond, this year.)

BT's timeline continues, seeing billion-transistor chips by 2009. And by 2010, they foresee that the highest paid entertainment star will be synthetic (think of today's popular synthetic Japanese star, Kyoko Date, with ten more years of technology under her virtual belt - http://www.kyokodate.com/ .)

DNA computers are foreseen by 2012, and by 2013 they suggest that "computer agents [will] start being thought of as colleagues, instead of tools." 2015 will see 3D videoconferencing, 2016 will bring to reality the Jetsons' "Rosie the household robot, " and in 2017, machine knowledge will exceed human knowledge. (I told you I hoped that all of these don't come to pass.)

2018, they suggest, will bring the 1-petabit memory chip. And 2020 will see "smart skin" that can help repair the fragile humans that still inhabit the planet. Incongruously, considering that I'm writing this at 37,000 feet while traveling at 560 miles per hour, they expect that the planes of 2020, although larger and with 6,000 miles of range, will still be subsonic (no one-hour sub-orbital flights to the other side of the globe.)

This series of BT articles continues by exploring the Ebusiness of 2020, when 80% of homes have Internet access, most appliances are Internet citizens, and "plug and play finally works." (Hey, it does have its problems, but even today plug and play is MUCH better than the pre-plug and play alternative!)

They expect that my COMDEX run-in with Aibo (http://www.compaq.com/rcfoc/20001127.html#_Toc499300003) will have led to robotic pets becoming common, where "cute, cuddly robotic pets provide extra amusement by ambushing the cat." (I hope we can turn that behavior off.)

Customized clothes, cut to fit our body-scanned images, will be common. And all of this Ebusiness will be lubricated by "seamless electronic commerce" using the global electronic currency we touched on earlier. Of course, as they describe (and I won't give away why), this may lead to "playground black markets," and to various forms of "barter economies" that have significant tax-collection implications. The role of banks will be considerably different than today, when they will have to be far more agile.

The Future?

There are quite a few more forecasts in this interesting, and perhaps a bit technologically controversial, series of articles. And you, like I, may not agree with all of them; indeed, we may hope that some results that ARE technologically possible, do not occur. But the ideas are still worth reading, and considering.

We know that many changes are going to occur over the next twenty years, driven by changes in societies, in economies, and by the rapidly changing face of computing. Consider that in 1971, a 4004 microprocessor contained 2,300 transistors. Today's Pentium 4 contains 42 million, and as we've learned earlier in this issue, we may be seeing 400 million transistor chips within 4 years. Imagine what the chips twenty years from now will be capable of!

Which is why I believe that the more we consider how even far-out ideas may intertwine and morph and play out, the better we'll be able to guide our future into one that we can, quite literally, live with.

The beginnings of massive changes.
Cris

The "Rapidly Changing Face of Computing" is a weekly technology journal providing insight, analysis and commentary on contemporary computing and the technologies that drive them.

The RCFoC is written by Jeffrey R. Harrow, Principal Member of Technical Staff with the Corporate Strategy Group of Compaq Computer Corporation.

RCFoC is available to the public at no charge via the World Wide Web at http://www.compaq.com/rcfoc

Last edited: Feb 27, 2001

5. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
And if you missed that link to BT in that last post then here is a quick look at a potential timeline.

BT technology timeline - towards life in 2020

Artificial intelligence and artificial life

Date expected

Chat show hosted by robot 2003
Computer agent personal shoppers 2003
First synthetic (but organic) life form 2003
Domestic appliances with remote (networked) intelligence 2003
Real time language translation 2004
Toys with network based intelligence 2004
Machine use of common sense inference 2005
Behaviour alarms based on human mistake mechanisms 2005
Computers that write most of their own software 2005
Intelligent robotic pets 2005
First artificial electronic life 2006
First organism brought back from extinction 2006
Domestic appliances with personality and talking head interface 2007
Systems to understand text and drawings (e.g. patent information) 2007
People have some virtual friends but don't know which ones 2007
AI models used extensively in business management 2010
Artificial nervous system for autonomous robots 2010
Highest paid star is synthetic 2010
Smart Barbie with personality chip and full sensory input 2010
Expert systems surpass human learning and logic abilities 2011
Most software written by machine 2011
Home manager computer 2011
Intelligent robots for unmanned plants 2012
Machine use of human-like memorising, recognising, learning 2012
Computer agents start being thought of as colleagues instead of tools 2013
Satellite location devices implanted into pets 2015
Office automation systems using functions similar to brain functions 2015
Machine use of human-like creativity 2015
Human knowledge exceeded by machine knowledge 2017
Electronic pets outnumber organic pets 2020
Electronic life form given basic rights 2020
Artificial insects and small animals with artificial brains 2020
Ubiquitous embedded intelligence 2020
Learning superseded by transparent interface to smart computer 2025
Living genetically engineered Furby (TM Tiger Electronics) 2040
Robots physically and mentally superior to humans 2050

Cris

7. ### Neujmin1Registered Member

Messages:
3
AI

There is no doubt that computers will or already do have superior memory and/or computing speed. But the thing is is that can we truely make a statement like that , that computers will be "above" humans, when we really don't know, still to this day, the true potential of the human mind?

8. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Neujmin1,

Yes I think we can because we can already largely determine the structure of the human brain and count the neural connections. The structures are really quite simple it is just the large number of connections that make the brain appear complex. We could model the human brain now and it would take an enormous number of present day computers, but the limitation of the model is that current computers simply do not run fast enough to mimic the speed of the human brain.

We have two primary problems: Physical size, and CPU speed. We can see that over the years CPU chips have become smaller and faster. The best estimates indicate that this trend is unlikely to change for some time and is likely to go beyond the limits for emulating the human brain.

The other obvious challenge is developing the software to replicate human thought processes. At the moment that is difficult to even test because of limits to computer speed, a problem that hit the early pioneers of AI many years ago when they massively underestimated the computer speeds they would need. But we are beginning to see progress on simple subsystems like speech recognition and generation. Vision is also gaining a great deal of attention and we will see complete solutions to that quite soon.

As each function of the brain is isolated and studied in detail then we will find ways to duplicate those functions in software, piece by piece, by piece.

In most people the full potential of their brains are never realized. As they obtain knowledge over the years and start to become productive, biological aging diseases kick in and start to erode those wonderful abilities. So far all people die and no one ever reaches their full potential. Machines with machine intelligence and unlimited life spans will not have the disadvantages of biological decay and may well achieve those thought potentials never reached by any human.

So it seems to me a crime that people have to die and lose those opportunities, but there is a solution to that as well: Mind uploading, and for that you should visit the Mind uploading forum on this BB.

Have fun
nd BTW welcome to the forum.

Cris

Last edited: Feb 28, 2001
9. ### Doc BrownRegistered Member

Messages:
20
Is it possible to duplicate the cerebral cortex in a robot? If not, we will have robots that are intelligent but not self-aware. If so, then all it takes is one robotic Hitler... [hums March of Death]

10. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Doc,

There was a discussion a little while ago where we debated self-awareness and intelligence. I think the general conclusion was that self-awareness becomes apparent at a certain degree of intelligence.

In other words self-awareness is inevitable as intelligence increases. From this we can conclude that if a robot did not have sufficient intelligence to be self-aware then he would not have the capabilities to become another Hitler.

I think what you were trying to imply was a sense of morality. That I think also comes from an appropriate degree of intelligence where the subject has the ability to reason and understand the implications of ones actions.

And that is a great topic – robot morality.

Cris

11. ### Tony H2oRegistered Senior Member

Messages:
441
Hi Cris,

Took a moment out from the relig discussion to get some thoughts from you.

The other night I was contemplating the very process of learning. Of how we learn, what we learn and why we learn the very things that make us human. I was considering this in the light of AI and what advances would be needed to develop AI. (at 3 in the morning)

You mentioned that you think as we separate each function of the brain that we will be able to essentially reverse engineer it to programing code.

I would disagree, I think it goes beyond just programing. It comes down to learning and how we learn.

Now supposing we had the hardware to mimic how we learn?

When I say hardware I'm referring not just the processing speed or the memory to match human brain function but the devices that feed that machine the information from its surrounding environment. The senses that we as humans use to learn. It seems to me that even if we could build a machine that duplicates how the brain functions we would also need to develop the sensors to feed that machine the information from its surroundings so that it could "learn".

We would then only need a simplified code that can deal with the inputs from those senses and react accordingly to limits set. ie. Maximum temperature the sensor can cope with before being fried. Or maximum pressure it could withstand before being crushed etc.

For us to develop AI that even comes close to human (if that is the objective) then we would need to develop a sensory organ's as complex as human skin, as human eye's, as human smell, taste and hearing. Devices like the tongue that can sense taste, temp, pressure, smell all at the same time. Without this technology I think that man's dreams of AI are just that. Dreams.

To make a machine that mimics human one must make a machine that learns through interaction with its surrounding environment. Learns through senses as complex and human senses.

I'm reminded at the moment of a scene out of Star Trek where Data has been captured by the Borg and is being given a human skin graft. The Borg Queen then blows gently over the graft and Data experiences senses that only humans can.

Do we really want to be the Borg? An assimilation of man and machine? All connected into the UWW (Universal Wide Web), all assimilated as one?

Just some thoughts.

Allcare

Tony H2o

12. ### Doc BrownRegistered Member

Messages:
20
I'm sorry, I guess I wasn't clear on what I meant by intelligence. What I meant was intelligence on the order of a dog or a cat. I don't know if it's possible to simulate the cerebral cortex, though.

13. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Hi Tony,

Yes nice to meet you somewhere else. And some good questions here.

I agree and I oversimplified my earlier statements.

There are two distinct steps to successful AI, at least within the limited context of this discussion. The first is the development of the processing algorithms and part of that depends on successful analysis of individual brain functions. The second step is producing the functional data on which the processes will operate. This data is distinct from input data.

Let’s use the eye for example. In AI I would have a vision algorithm (set of computer instructions) that will use a lookup list (functional data) to tell it how to interpret the incoming data stream from the vision input device.

At the outset the functional data would be missing. A learning process that essentially adds value to the basic vision algorithm would generate the data. The learning mechanisms use feedback loops, which help to continually refine and perfect the functional data.

The problems AI researchers have found with this type of neural network learning feedback loop is how to determine when to stop. Sounds easy right, but the learning process doesn’t know when it has done enough and continues to iterate and can actually make things worse. However, these are just matters that simply require more research. There is nothing inherently wrong with the fundamental concept.

Now if we assume we have a complete robot with all its peripherals and input devices hooked up, and all algorithms loaded, and then we turn it on. Remember at this point it has no functional data (i.e. it hasn’t learnt anything yet). The machine will essentially have the same capabilities as a newborn human baby. And the learning process could conceivably take almost as long as it would for a child to reach adulthood.

The big prize comes when the robot has acquired significant functional data that we would consider young adulthood and then we can simply take a memory dump and create clones. Subsequent robots would never have to go through the basic learning steps again. But note that the learning processes would continue to function and the robots would become cleverer the more they interact.

But that scenario is unlikely since the development of subsystems, like vision, will be developed independently together with appropriate functional data. When the final robot is assembled and switched on all its subsystems will already be in a mature state. Full adulthood would only take a short time if any.

A word of warning: I used vision as an example of a subsystem, and it is a true subsystem, but it is also difficult to separate the process of detecting and capturing detailed images and interpreting those images. For example if you catch a partial glimpse of the back of someone’s head there might just be enough information there for you to project that data and complete what is missing and be able to recognize the whole person. That is a combination of partial image capture, recognizing that the image is part of something else, and then a database search on that partial match. We can conceive of the entire sequence and have achieved much of this in the laboratory. We just need more computing power to process more complex images and faster. I work for Compaq as a software designer, I know that our Cambridge labs in Massachusetts have made significant progress in the area of vision interpretation. It is not science fiction.

I think you conceive the human form as something near perfection. Whereas it is really a very poor design. I can quote you a long list of design failures if you wish, but for now consider: Eyesight (corrective lenses), vision (limited to a tiny part of the electromagnetic spectrum), teeth (more than half the world population do not have their own teeth), metabolism (a disaster in energy transfer engineering), touch (extremely slow signal propagation). Have you ever burnt your finger because you touched something hot? If your touch sensor and brain could have been faster you would have detected the heat before touching it.

Not only will AI and robotics come close to humans but will far exceed the human design in every aspect and within the next 50 years if not very much sooner.

It will be worth starting another thread to explore the research into artificial skin and other robotic sensors.

There are significant problems with the beguiling and compelling illusions manufactured by the modern movie industry, especially in the area of science fiction. These are images generated by imaginative artists and although they give minimal credence to the real world of science, the stories remain imaginative fiction. So be careful not to confuse illusions with what is or might be possible.

And I’ll give you an example: Have you ever read any of the Dragon books by Ann McCaffery? (spelling might not be quite right). These are wonderful stories of a planet where a nearby planet passes every few centuries and sends poisonous spores that decimate all living things. The people form an very amicable alliance with fire breathing telepathic flying reptiles (dragons) and together they burn the deadly spores in the air. Before Ann’s stories, legends always portrayed dragons as evil creatures that needed to be slain, in her stories dragons are majestic and empathic creatures. So you see that a horror story written by one writer can be reformulated into the opposite by another artist.

The Data of Star Trek is not how my current knowledge leads me to envision robots of the near to mid-term future. Emotions and feelings appear to be generated by just another part of the brain. Duplication should be no more difficult than any other subsystem. And that is not to imply that any part will be easy, but we do need those faster processors before we can start making real progress.

The Borg is portrayed as an evil concept in that individuality is lost in favor of a collective single mind. Now consider just one simple change, imagine that the drones had individuality and voluntarily joined with all the others when major decisions had to be made that would affect all of them. When thoughts can be communicated at light speed as opposed to painfully slow written or verbal communication then one can see how decisions could be made very rapidly. A final vote could be completed in seconds.

I very much would like to discard my fragile and vulnerable poorly evolved biological shell in favor of a superior non-biological version. And if bio-electro-mechanical augmentation becomes available that give me access to massively enhanced senses then I’ll be the first on the list.

So we have disagreed as usual, but I hope I have given you some new ways of thinking about AI and robotics.

Take care
Cris

14. ### Tony H2oRegistered Senior Member

Messages:
441
Cris,

Thank you for your detailed response, I appreciate the time you took to compile it.

Everything you said sounds extremely plausible, although our best visions of the future always manage to fall short of the reality of it.

However I find I am left with one thought spinning around in my head, "unnatural".

Let me clarify this.

Lets say for a moment that we do manage to achieve this goal of AI and even beyond that the possibility of humans becoming a totally non biological mechanism through memory / emotion dump.

What have we then become in comparison to the environment that surrounds us?

A machine with a significant advantage over biological life forms, A device that does not need to rely on the surrounding bio systems for support of its operation a device that could ultimately destroy the surrounding biological system.

Now I'm not trying to paint an evil picture of this because I am in a way playing the devil's advocate by taking an opposing view to the rosy picture you painted of this technology.

So the real question is how would you or others see this technology interacting with biological life forms (again scenes from the Terminator, the Borg, Blade Runner, Number 5 is alive.....etc). Especially when you consider that the created entity would have a "mind of it's own" and may not want to yield to inferior human rules and regulations. Especially when you consider that this "lifeform" with it's advanced memory recall, data processing speed, sensory array, who perhaps would consider itself superior to human life.

If we think of this in the light of "evolutionary" thinking, this entity would have made a significant jump on the "natural" process of "evolution". Not being driven by necessity to "evolve" to suit and interact with the surrounding conditions and environment, but created to exploit them? (strong term). Created to overcome the fear of the race that created them? The fear of death and mortality?

I don't know what others think but to me that sounds.............well unnatural is the best way to describe it. As marvellous and incredible as this technology may or may not present itself I think it is still important to look into how it will not just affect us but also the world in which we are but a part.

Technology without reason or wisdom could be dangerous.

No line between man and machine? Maybe there should be?

Allcare

Tony H2o

15. ### BorisSenior MemberRegistered Senior Member

Messages:
1,052
I'm late to join a forum I should've joined long ago; apologies, the excuse -- time constraints.

As usual, a few comments to start:

Software is indeed becoming the limiting factor in AI. There appears to be little doubt that enough computational power will be available in a quarter century to emulate a human brain in real time. However, I doubt there will be software sophisticated enough to do that. For example, right now we have massively parallel supercomputers capable of a petaflops (around 1,000,000 times faster than a typical PC) -- but they and their lesser cousins run hydrodynamics simulations or genetic sequencing algorithms, not AI. That's not incidental. Part of the reason is of course pragmatic, but think of the advantages the military could have with an intelligent missile as smart as a trained homing pigeon. The real reason is that nobody has the theoretical or empirical underpinnings to emulate even anything as "simple" as a pigeon, much less a human. The task of mapping and mathematically unraveling the brain is much more arduous than construction of fast computers, and is still in its infancy. I want to be optimistic for the short-term, but having personally dealt with the enormous difficulties of biological and machine vision, I am not.

Thing is, brute power is not always the answer; without the right algorithms certain problems remain intractable, or are at best solvable in only mediocre fashion. Case in point: chess. It took Deep Blue with some outrageous processing capacity and an enormous database dedicated specifically to a particular task to beat a sophisticated human whose brain nevertheless clearly wasn't even designed for that task, and whose total chess knowledge, though doubless impressive, couldn't have equalled the database. Another case in point: speech recognition. Current solutions approach "perfection" as defined by the human standard, but will never reach it regardless of how much machine power is thrown at them because they are algorithmically flawed. And it is very hard to discover the right algorithms, because that task usually requires first an understanding of the problem... Well, anyway, enough doom and gloom.

As a separate issue, I don't entirely agree with Cris' assessment of the human body. I don't see it as somehow primitive and crude. While machines can be repaired with spare parts and maintained externally, the human body repairs and maintains itself. That's a capability machines will never have, unless they become built from near-autonomous mutually redundant building blocks analogous to human cells.

Personally, I envision robotic perfection as mechanisms built out of micromachines (in all likelihood quite a bit larger than Borg nanoprobes in Star Trek lore), which would make the artificial lifeforms self-generating and self-repairing, far more resilient and autonomous in the long term. Such component-wise construction could make possible real-time shape morphing, making Transformers look like a joke.

And I agree with Cris that shifting toward artificial life will bestow a wonderful new universe upon us, essentially giving us an eyesight as if before we were blind, giving us legs as if before we could only crawl. Imagine seeing across the entire electromagnetic spectrum. Imagine feeling the lines of magnetic force twist and dance through space, creating photonic displays. Imagine being able to pick up scents a million times better than the best hound. Imagine sensing an earthquake a thousand miles away. Perceiving the sounds of magma underneath the continental crust, all the way to the sounds of meteorites zipping through the stratosphere. Telescopic and microscopic vision without the need for either external instrument. Flawless sense of orientation and direction, velocity, acceleration. The subtlety of lifting an ant by its leg, and the strength to dig through solid granite. Temperature, humidity, acidity, pressure, radiation -- you name it and the sensors for it could be all over your body and tied directly into your sensorium. Telepathy made real. Imagine transforming into a flying "bird" and soaring up through the clouds, coming down through the ocean's surface, transforming into a "fish" zooming along the ocean floor. Thinking and perceiving so fast that a speeding bullet slows to a crawl, and you catch it by gently applying force so as to not rip the bullet apart. Being able to hover above the plane of the ecliptic and to slow down the perception of time so much that planets zip around the sun on their elliptical orbits like bees around a flower. Being able to hold an encyclopedia's worth of data in short term memory. Being able to focus attention of thousands, even billions of things simultaneously. Carrying on thousands of concurrent conversations with as many different people. Being able to literally see through someone elses eyes, or to literally share your experiences -- with no need to first translate them into words. Practically eternal life. Such freedom as you may only imagine in your dreams, and more. None of it can be thought of as human or somehow dehumanizing -- only superhuman and incredibly liberating and empowering. A real-world version of Biblical paradize, on steroids.

Although I doubt any degree of technological sophistication could equally maximize a machine's performance in deep space and deep ocean, in the sky, and on the ground, in seas of liquid methane and in a lava flow, yet enabling it to be flexible and versatile enough to provide more mobility and functionality than a human body in all of those situations. The performance characteristics are just too different. Perhaps mind uploading would make it possible to change from one body to another at will, "adapting" to different environments. Another intriguing possibility is travel at lightspeed, so to say, by transmitting the mind on a laser beam to a distant location where it is transferred into a local body.

As for natural vs. unnatural, who is to say which is which? Yes, it is entirely possible that human-made life will overcome and exterminate pre-human life including humans as well. However, nothing could completely exterminate all biological life on Earth without in effect destroying the entire planet -- which not even a mad robot would consider practical or worthwhile. But regardless, should such an abrupt shift from one form of life to another occur, it can be thought as merely an evolutionary leap. 1,000,000,000 years in the future from that event, it will only be another signpost on the road of natural history. After all, who but the orthodox theists could claim that deliberately designed life is not the ultimate reason for existence of, or the next inevitable or logical step in a long series of steps after, biochemically evolved life? Who is to say that all intelligent life in the universe (should there be any) does not eventually evolve along the lines of self-design? The universe could be full of marvellously intricate and intelligent sentient robots -- could that be a reason why we haven't been visited by an extraterrestrial intelligence? After all, they could consider us merely as precursors to what they view as real life -- and who wants to engage in intellectual exchanges with primordial slime? But this post is beginning to sound like it belongs in the UFO forum, so enough of that.

Last edited: Mar 2, 2001
16. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Boris,

I’m very pleased to see your views here; you always provide a good read with real value. You’ve also touched on quite a few areas that I hope to explore more fully later on, providing there are sufficient contributors, or maybe I’ll just post anyway. Simply the discipline of marshalling my thoughts to write a decent post is a worthwhile activity in its own right.

You mention Deep Blue: Did you know that that machine did not use any AI techniques? It was simply massive computing power, sheer brute force with no real subtleties. I understand that Karpov became quite distressed because he felt he could detect subtle strategies that he had only ever seen before in humans. He was convinced that there were humans helping the machine. He had played against so many computers before that he was sure he could recognize a machine player because of its simplistic approach. So perhaps super parallelism combined with super power might just provide enough brute force to emulate intelligence without actually being AI, hmm, in which case is there a difference? Nah, we can do better.

Self-repairing robot bodies sound great but I see that as just a copy of the human paradigm. We are free here to create a simpler, cheaper, and more flexible approach. Self-repair requires significant sophistication and complexity and is likely to detract from advantages such as low power consumption and additional sensors. And self-repair in humans isn’t so great when you lose a major component, e.g. an eye, leg, arm, etc. In other words if you lose something really important you are still dependent on an external repair service, doctors for humans, and mechanics for robots. I would go for simplicity and build integral strength from the outset that largely negates the need for the self repair system, e.g. a steel arm will be more resilient than one of flesh and blood. I work with massively parallel fault tolerant systems, and it is this tactic that I would build into robots. Humans are largely fault tolerant when you consider that you can lose one of something and still survive, e.g. lose one eye but you can still see, lose a kidney, lose an arm, etc. Even the loss of a brain hemisphere, when young, can still result in full survival. I would build robots with greater fault tolerance that would enable them to survive long enough to reach a repair shop. Why have limits of only two of everything, four eyes (two front and two rear) could be pretty neat. Aesthetics wouldn’t be a primary concern.

Moving further into the future I had imagined a processing element perhaps the size of a sugar cube that could generate some form of shaped force fields that could give the illusion of shape and texture. Perhaps like the holograms of Star Trek fame. This would allow deep ocean and space voyages without changing bodies. But this is just fantasy – fun huh?

I like your final statements: They are identical in context to a post I made in one of the alien forums last year, “The Reason Why We have Not Been Contacted Yet”. I wouldn’t be surprised, if there are any aliens that they are simply waiting for us to mature to their level to allow meaningful communication. They would possibly see us now, as we would view a planet of apes, i.e. there would be little reason to make contact with low intelligent apes.

It seems there are two viewpoints concerning the progress needed in neuroscience that would allow us to make significant progress in software. I’ve been following that debate in the MURG yahoo.groups forum. I understand your position, but I remain more optimistic, although I do recognize you are closer to the nitty gritty than me.

Cheers
Cris

17. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Tony,

Your last post has some important points, with which I am in total agreement, and you are not alone with those concerns. But I want to respond with some more detail, hopefully tomorrow.

Cheers
Cris

18. ### BorisSenior MemberRegistered Senior Member

Messages:
1,052
Cris,

Fields, huh? Yeah....mmmmm back to reality.

It is fun though, true that.

Actually, I'm not concerned with complexity of self-repairing shape-morphing self-constructing robots with a cherry on top, since they are going to be designed by already-enhanced humans or simpler robots, not by us simpletons. I don't propose that my vision be the immediate goal, rather I see it as a realistic <i>eventual</i> achievement some time in the future, at least knowing what we know now about the universe (and assuming we aren't going to discover the Schwartz, admittedly a poor assumption.) But if indeed there is that much more about the macroscopic universe that we still don't know, then everything I said can be taken a million times over, and with fries too.

But in defence of my modular proposal, actually it is the ultimate embodiment of redundancy: every constituent part carries in itself the capability of generating the rest of the body even if it's the only remaining part after some big bang (much more than the structurally specialized cells of the human body; these things are specialized only functionally, and even that's reconfigurable at will.) This isn't so much about personal survival as it is about survival of the species and perpetuation of life in the universe. In this case, computation can be distributed throughout the body, and instead of a sugar cube you have a reconfigurable cluster of a billion sugar grains. A little microcosm all of itself, in other words (cool if not the most efficient, but hey, even now the world's moving toward cluster computing.) O hell, we're brobably both sounding like utmost baffoons anyway, so never mind them little things

19. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Boris,

Ah much too much imagination. I think maybe we should concentrate more on the near future and determine what will become reality. I know, kinda boring compared to fantasy land, sigh.

I can see that not only can you outthink me on conventional matters but I detect that you can out 'imagine' me as well.

Great stuff.

Take care
Cris

20. ### ATHEISTHATERRegistered Member

Messages:
25
Future of Artificial Intellegence (Silicon vs Bio)

It seems to me that silicon based advancements in computer science will continue rapidly during the next few odd years. In fact I think that the present down turn of nasdac will spur a competitive spirit of innovation. Hell,....err..heck, "ANY DAY NOW" I would not be surprised to hear some talking head (NEWS ANCHOR)announce a "major" advancement in computer technology. (Not a prophecy people, just an observation based on past historical observations).

I Certainly do expect that the gap between gentech (genetic technology) & AI to decrease rapidly. Will there be "Sci-fi" type Cyborgs among us? Yes. At least prior to the inevitable transition of technology from mechanical engineering to bio engineering. In my opinion this will depend largely on the relative advancements (race) among the silicon engineers vs the bio engineers.

??????Will there be silicon chip brain enhancements which could be implanted into a human being which will give a person the ability to Remotely choose to access the net, get & send email, trade stocks, play cyber games, hold interactive real time communications with others, and even download the ability to speak a foreign language simply by thinking about, and thus activate an implanted "chip"???? Yes, I am very sure that this will eventually come to pass.

As far as future unborn generations of human beings, I do expect that at some point genetic technology will become so advanced that in a hundred years (possibly a lot less), people will be "designed" in such a way that the implanted chips that we would love to have now will be considered to be more of a "crutch", for the mentally deficient underlings of the 22nd Century!

Regarding the evolution of internet communications, I expect sooner or later CRIS, that some rebellious, break all the rules, and generally pain in the ass student of science <i><b>("I Resemble That Remark" (Garfield the Comic Cat))</i></b>will prove the existence of telepathy and be a worthwhile contributor to the one day common ability of humans to communicate freely without this most primitive, silicon based internet that we all struggle with today. That will be a day worth living for,,,,,,,,,,,,,,,,,,,,*&!$*&!$@^!\$# err.........??????worth being reborn for/////////////

thanks______

Last edited: Mar 9, 2001
21. ### CrisIn search of ImmortalityValued Senior Member

Messages:
9,188
Thanks AT, great post. Good to see you here.

22. ### ATHEISTHATERRegistered Member

Messages:
25
Thank you sir. I appreciate your kind words of encouragement.

<i>Having said that, I feel the need to "tear loose" and write a mass of incoherent and incomprehensible thoughts...............on the subject........</i>

??????????................

Perhaps another time.

<i>" This town needs an enema" <b>(The Joker)</b></i>

Last edited: Mar 9, 2001
23. ### ATHEISTHATERRegistered Member

Messages:
25
YA know Cris, the continuing evolution of the human mind is an absolute and raw reality of the "succeed or die" rationale which happens to be a simple yet complex law that governs not only biology, but the entire universe, as I have visualized it over a past many years.

The evolution of AI has obviously had it's parental helping hand and for a while longer at least, will continue to be nurtured by it's mother (human software designers).

But one day soon Cris, HAL & it's bretheren will eventually leave the nest. How this child of great collective power will behave will have a lot to do with how it's parent (todays software designers) RAISE IT!

In my opinion, Hg Wells that I purport to be, this son of man can become a wonderfully helping hand to man, but so far as fear, methinks that ultimately it will be the primal emotional drives of greed, within the human mammalian brain that might possibly cause the AI's(terminator/HAL) to outgrow and become aggravated with all of us.

Personally I don't have a problem with AI . At least so far "computers" seem be laid out firmly on the basis of logical rather than emotional drives. It would seem to me that the procreage of the current generations would continue to be helpful to the human race for a long time to come.

After all, that's the initial idea of AI. To be helpful to humans who are presently unable to perform a variety of tasks. However if the very things (emotions), which have made all life successful are programmed into AI (a possibly necessary step), then it does become foreseeable that AI collectively may become a death knell to humanity. This is a highly doubtful scernario though.

CHEERS,

<i>"I have wasted my life" (Leonardo Da Vinci)</i>

Last edited: Mar 9, 2001