249 replies to this topic

[treonsverdery]

this page now at geocities.com/treonbarleyverdery

Edited by treonsverdery, 01 November 2006 - 12:28 AM.

[Bruce Klein]

Bringing the connection between AGI and Life Extension a bit closer to home, Novamente's sister company, BioMind LLC, recently helped the largest study yet of chronic-fatigue syndrome (CFS). BioMind uses Novamente AI technology. BioMind sifted through massive amounts of data in order to find the genetic link for CFS, which had thus far been mainly thought of as a psychosomatic problem.

CFS is estimated to affect one million people in the United States alone. Here's an excerpt from the Nature article:

Chronic fatigue has genetic roots
Massive data-crunch points to basis of inscrutable disease.

Researchers at the US Centers for Disease Control and Prevention (CDC) in Atlanta tackled the problem in a new way. They handed four teams of scientists a massive set of information about the symptoms and biology of CFS patients, and challenged them to pull out anything that might explain the disease.

One study showed that patients with CFS tend to have a characteristic set of changes in 12 genes that help the body respond to stress. They showed that a particular combination of gene sequences could predict whether a patient had CFS with over 75% accuracy.

Ben was quoted:

"CFS is a real bodily dysfunction," says Ben Goertzel of Biomind, a biotechnology company based in Rockville, Maryland, who led one of the groups. "The idea that these people are just tired is pretty clearly refuted by this batch of results."

[Bruce Klein]

ImmInst members may not be aware, but the inventor of the word "cyborg", Dr. Manfred Clynes, contributed a chapter to ImmInst's first book, along with two other AI pioneers, Ray Kurzweil and Marvin Minsky.

[Bruce Klein]

Harold,

I was hopping that you'd engage me a bit more. I'm quite impressed with your insightful questions. Thus, in an attempt to bring you back into these discussions, may I ask you a question? Do you see a Technological Singularity happening anytime soon?

[MichaelAnissimov]

Based on being somewhat familiar with its founder, its academic hosting and support, and an explicit AGI focus, I think that Mindmakers should definitely be added to the list of AGI projects.

[Bruce Klein]

Thanks Richard and Michael,

I've added both MindMakers.org and A-i.com to the list of AGI System Building Projects.

[Bruce Klein]

Highlighting more of the connection between AI and Life Extension, note Ben's work via BioMind on Parkinson's disease and mitochondrial DNA (ImmInst Topic) which has been a collaboration with Drs. Davis Parker and Rafal Smigrodzki.


http://www.biomind.com/pd_article.pdf

[Bruce Klein]

Typically people abstract the second part out of the equation. Why does that happen, exactly? The history of failure for these things, the perceived lack of availability of potential solutions, the technical depth required to evaluate such claims, lack of concrete information about claimed projects ...?

Right, and Peter Voss says it well:

"A great number of researchers reject the validity or importance of 'general intelligence'. For many, controversies in psychology (such as those stoked by The Bell Curve) make this an unpopular, if not taboo subject. Others, conditioned by decades of domain-specific work, simply do not see the benefits of Seed AI - solving the problems only once.

Of those that do not in principle object to general intelligence, many don't believe that AGI is possible - in their life-time, or ever. Some hold this position because they themselves tried and failed 'in their youth'. Others believe that AGI is not the best approach to achieving 'AI', or are at a total loss on how to go about it. Very few researchers have actually studied the problem from our (the general intelligence/ Seed AI) perspective. Some are actually trying to reverse-engineer the brain - one function at a time. There are also those who have moral objections, or who are afraid of it.

Of course, a great many are so focused on particular, narrow aspects of intelligence that they simply don't get around to looking at the big picture - they leave it to others to make it happen. It is also important to note that there are often strong financial and institutional pressures to pursue specialized AI.

All of the above combine to create a dynamic where Real AI is not 'fashionable' - getting little respect, funding, and support - further reducing the number of people drawn into it!

These should be more than enough reasons to account for the dearth of AGI progress. But it gets worse. Researchers actually trying to build AGI systems are further hampered by a myriad of misconceptions, poor choices, and lack of resources (funding and research)."

http://www.adaptivea...earch/index.htm

[Athanasios]

I have been reading this thread, and am liking the AI that novamente has pumped out so far. The narrow AI has been shown to be very productive. If through pattern recognition, and integrated databases, the AGI system can utilize the narrow AI, it will be a huge boon. It doesnt have to be as "smart" as a human nor "conscious" to achieve this, IMO. Although, the editing of its own code and whatnot seems to be very ambitious, and may or may not require the consciousness aspect, I dont think having it in the timeline is detrimental to the cause of the project.

[Bruce Klein]

Thanks, cnorwood19.

Our next focus for NovaBaby is to virtually emulate the gyrational characteristics of this model so that our future humanoid robotics will be more tactile. Should be fun...

[Athanasios]

Should be fun indeed... hahha

Dont alow Novababy to wear that much makeup though

[Bruce Klein]

Do I see TR happening any time soon? Until such time that we can simulate the level of complexity that involves the development of a cell to an organism, no I don't, particularly when at present we have difficulty predicting the folding of a single protein.

This last sentence hints at why we need AGI. Finding patterns in large amounts of data is what is required to understand the protein folding problem... a task more suited to an artificial scientist than a human one.

[Bruce Klein]

What we do know with a high degree of confidence is that intelligence is an emergent property. Take a genome, place it into a cell, add the right environment and you can create a Mozart or an Einstein. Now that is more extraordinary than any contemplation of a singularity.

Biological life is quite fascinating. However, this should not mean that the prospect of artificial life is to be discredited only because it is in development... especially when it has the potential to solve so many problems, including aging.

[Bruce Klein]

Yet in 2005 glia (another cell type found in the brain in addition to neurons) were found to contribute to neural networks in a way that had never been understood before. There are likely further levels of regulation and encoding of information that may act on even more subtle and as yet undiscovered levels. The human brain still remains the most complex object known to man in the universe. Therefore I am very doubtful of those who propose that a "mind" may emerge based on their knowledge (or more likely lack of) of neural networks. If one is seeking to model intelligence one must understand the principles - the rules - by which it works.

As hankconn replied, and I have already replied earlier here, what happens if we map the brain and then find out that it's a complete mess, and thus impossible to scale to greater-than-human level intelligence? Then, we'll have to go back to the drawing boards and find a more non-biologically inspired approach, no? With this in mind, Novamente is taking a shortcut. This graphic illustrates this point:

Approaches and Projected Time Frames in Reaching Artificial General Intelligence (AGI)

As knowledge of the human brain increases, and the cost of computing power decreases, more scientists understand how creating powerful Artificial General Intelligence (AGI) via emulating the human brain in software is possible.

Currently, however, a substantial knowledge gap exists between our understanding of the lower-level neuronal mechanisms of the brain, and our understanding of its higher-level dynamics and cognitive functions. Creating AGI based on brain mapping must wait until quantitative improvements in brain scanning and modeling lead to revolutionary new insights into brain dynamics, filling in the knowledge gap. There is little doubt that this will happen, but it is hard to project how long it will take. Kurzweil estimates 2045[1] based on systematic extrapolation of the observed rate of improvement of brain scanning technology.

Computer science based approaches to AGI, on the other hand, provide an exciting possible shortcut. There is no need to wait for brain scanning to improve and neuroscience to undergo a revolution, leading AI theorists such as Marvin Minsky[2] agree that, with the right AGI design, contemporary computing hardware is quite likely adequate to support the implementation of AGI at the human level and beyond.

Skeptics will point out that the computer science based approach to AGI has been pursued for some time without dramatic successes. But computers have never been as powerful as they are now, and more importantly, the field has lacked adequate AGI designs, taking into account the comprehensive knowledge gained by the fields of cognitive and computer science as well as neuroscience.

If pursued properly based on a powerful AGI design, the computer science approach to AGI may be able to lead to AGI at the human level and beyond within the next decade. And these computer science based AGIs will then, among other transformative effects, drastically increase the rate of progress of science and engineering, including brain mapping and neuroscience.

1. Pg. 136 "The Singularity is Near"
2. http://www.novamente...file/AAAI06.pdf

[Richard Leis]

Quote prometheus:

There are likely further levels of regulation and encoding of information that may act on even more subtle and as yet undiscovered levels.

How do you known where we are in our knowledge about brains and intelligence? Determining where we are is in fact an important part of scientific research. Sometimes we do not necessarily want to know how to do something, but if that something is even possible to do. However, I am not certain that your statement offers anything concrete by which to debate Bruce.

I am not going to go out on a limb and tell you that you must believe Bruce and that Novamente will absolutely succeed in their task because they have covered all their bases. I am also not going to outright dismiss the effort with vague statements about how there likely is something they have missed.

[Richard Leis]

Back in 1991 or 1992, one of my coworkers, a computer programmer, was asked by other programmers to take a look at what Tim Berners-Lee was working on. They were playing around with HTML and a web browser. He immediately dismissed it as hogwash and said it was not going to amount to anything.

Last year I asked another coworker, also a programmer, what he thought about the web service push and the use of AJAX and other technologies to build applications accessed via a browser. He insisted that he certainly did not want to access programs via the web and that the effort was not going to be successful. Whether or not it has been successful may be open to some debate, but the amount of effort pouring into the endeavor has grown unabated and I personally have grown fond of access to these programs and my work via any computer I use around the country.

Is this particular thread to dissolve into a dismissal of AGI as too hard, with no bearing on Life Extension?

[MichaelAnissimov]

Artificial Hippocampus

From the above article:

No one understands how the hippocampus encodes information. So the team simply copied its behaviour.

We regularly simulate things without understanding how they work. If we can simulate the hippocampus, we can simulate the whole brain. Consciousness is merely what the brain does with its physical structure - we need not go to mountaintop and meditate on the deeper meaning of consciousness and achieve enlightenment in order to simulate it. (Some articles on AI talk as if that would be necessary.) There is nothing special, nothing fundamentally different about our brains that gives us consciousness. It's probably just a threshold quantity of self-watching feedback loops.

Also, have you seen comparisons of the computational power of the human brain to contemporary supercomputers? These studies strongly suggests that we are in fact very close to possessing the technology to implement human intelligence.

And let's go even further and say I have a scanning device that can read the structural content of the human brain on the molecular level, then I go simulate that in a nanocomputer of the future. Why must I understand what I am copying? Where does the strategy go wrong? It only seems like an implausible approach if you abandon functionalism (things working by virtue of their interactions) and embrace substrate essentialism (there's a special sauce in carbon that gives us think-power).

[MichaelAnissimov]

For others it is a yearning which is so deep seated that I am beginning to think of it as a transhumanist replacement for religion.

Who? Can you name names and cite quotes, even if you need to do so privately? I've spent extended time talking with or hanging out with the core players at SIAI, and remarkably enough, they all act and talk like pretty normal (non-religious) people. Have you ever talked to an engineer trying to design a huge office building for the first time? Or a biologist attempting to achieve an important result? Or a programmer trying to get a new business application out the door? Nowadays, everyone is excited about their focus in life, and these accusations of religiosity are pointed at nothing but straw men - these people plainly don't exist.

What creates the association with religion is that AI is an attempt to build something more powerful than us. But instantly associating the effort to create something more powerful with the idea of worshipping something all-powerful and non-existent is a side effect of mankind's messed up memetic past, and has nothing to do with legitimate present-day efforts towards both narrow and general AI.

[John Schloendorn]

Why must I understand what I am copying?

Makes me wonder if there's any way an aged upload can continue dying from Alzheimer's because nobody knows how to cure it even in silico... Probably they'd just freeze him until they figure it out somehow -- Argh, we need that world, it would make things so easy :-)

[Bruce Klein]

AGI is in the same spot powered flight was about 120 yrs ago. Ask most anyone then, scientists especially, and they would have said it was a far-off nearly impossible dream... close to a religiously inspired dream at that, seeming that only gods could fly. Hopefully, as MA suggests, we needn't try and convince everyone in order to make it a reality.

[jaydfox]

Hopefully, as MA suggests, we needn't try and convince everyone in order to make it a reality.

I'm not really interested in convincing everybody. I'm not even interesting in convincing our resident critics that it's possible.

I'm mainly interested in making it clear that it's not "obviously" impossible, and should remain something ImmInst advocates, because it is relevant, even if we disagree on how important it will be in the short-, mid-, and long-term.

And when I say it should remain something ImmInst advocates, I don't just mean advocacy by individual members, but as something the institute itself supports, the way we support life extension via biotech (e.g., SENS). Whether that's through a declaration of principle, or space in our next book, or a panel at a future conference, etc., it should be something we support officially. If our resident critics have a problem with it, then debate it in the fora, but don't try to have it removed from the institute's official focus.

Edit: changed "only" to "mainly". I mean, I suppose I wouldn't be upset if they changed their minds...

Edited by jaydfox, 07 July 2006 - 01:25 AM.

[jaydfox]

Hopefully, as MA suggests, we needn't try and convince everyone in order to make it a reality.

I'm not really interested in convincing everybody. I'm not even interesting in convincing our resident critics that it's possible.

I'm only interested in making it clear that it's not "obviously" impossible, and should remain something ImmInst advocates, because it is relevant, even if we disagree on how important it will be in the short-, mid-, and long-term.

And when I say it should remain something ImmInst advocates, I don't just mean advocacy by individual members, but as something the institute itself supports, the way we support life extension via biotech (e.g., SENS). Whether that's through a declaration of principle, or space in our next book, or a panel at a future conference, etc., it should be something we support officially. If our resident critics have a problem with it, then debate it in the fora, but don't try to have it removed from the institute's official focus.

Cross-post:
http://www.imminst.o...00

Hmm, Harold, I think I you an apology. In the chaos of ideas being tossed around, dropped, morphed, etc., and the incendiary exchanges here and there, I at some point seem to recall that you shot down Bruce's attempt to have AGI and other technologies given a similar limelight to biotech.

Skimming back, I see now that, quite the opposite, you even advocated DoPs for cryogenics (cryonics), transhumanism, and the singularity.

Now I'm confused. Until I figure out where I got so far off base, please accept my apology.

[Bruce Klein]

Topical to this discussion, I've just uploaded Ben's AGI Workshop (sponsored by Novamente) introduction where he talks about how he got into AI and the current feeling within the scientific community about AGI.

10 minutes

http://www.agiri.org/ben_agi.htm

[jaydfox]

1. We have no practical definition of consiousness, therefore we cannot model it.

AGI doesn't have to be "conscious", in whatever mystical or functional way we might be talking about (since I'm sure we all mean different things by "conscious"), to be intelligent enough to self-improve and scale with increased computing power to something far more intelligent than humans. Consciousness is a sufficient, but not necessarily necessary, condition, for human-level intelligence in the domains required to spark a singularity.