RoR: Forum

Post 0

Thursday, March 23, 2006 - 6:12am

Link

Edit

Thanks for the review! Definitely sounds like Kurzweil's exhuberant outlook is continuing, and that this book too will be fun to read. It sounds like he's touting things that are probably technically possible - but trying to extrapolate an exponential trend to determine timeline is never a safe bet.

I really like your comments on singularity and Objectivism. I think the problem of knowledge transfer being so cheap is one we're already seeing in smaller form now, with intellectual property concepts needing to evolve as information becomes more and more clearly a non-rival good. A non-biological entity being conscious? - Absolutely! (and the bloody system of you, the book and the room know Chinese too! :) )

The virtue of productivity, political activism and virtual reality ones I've not encountered before, and are excellent questions. Something to think about.

Post 1

Thursday, March 23, 2006 - 8:32am

Link

Edit

There is a vital economic imperative to create more intelligent technology. [...] We will continue to build more powerful computational mechanisms because it creates enormous value. We will reverse-engineer the human brain not simply because it is our destiny, but because there is valuable information to be found there that will provide insights in building more intelligent (and more valuable) machines. We would have to repeal capitalism and every visage of economic competition to stop this progression.

-Ray Kurzweil, "The Law of Accelerating Returns," 2001

I like this guy already.

Post 2

Thursday, March 23, 2006 - 9:21am

Link

Edit

Thanks for the excellent review. I have been an avid reader of Kurzweil for sometime and have been interested in extropian and transhumanist ideas before I was interested in objectivism. To me these philosophies are not very different, although extropianism and transhumanism lack philosophical direction, they and objectivism all embrace individual life as the core value. Some other people have written on uniting these seperate idealogies and I have mentioned this in some previous threads. The following article is a decent one of the subject.

The Objectivist-Extropian Synthesis
http://www.geocities.com/rational_argumentator/OEsynthesis.html

Also I recommend checking out Ray Kurzweil's online essay

"The Law of Accelerating Returns"
http://www.kurzweilai.net/articles/art0134.html?printable=1

In regards to your review, and your questions.

If the primary choice of every living thing is to live or not, what is the primary choice if man eliminates death?

that implies that the only purpose of life is to struggle against death, and without the threat of death life has no meaning. The struggle against death is the first purpose of life, once complete, we can move on to other things. The primary choice then becomes to progress or regress.

If all knowledge can be transferred and learned with no effort, what happens to the virtue of gaining knowledge?

The information still has to be processed and an attempt made to understand it, just as it is today. You can have perfect memory of something but that does not mean you understand it and integrate it into your life and actions completely.

Michael F Dickey

Post 3

Thursday, March 23, 2006 - 5:21pm

Link

Edit

* If the primary choice of every living thing is to live or not, what is the primary choice if man eliminates death?
There is no such thing as elimination of death. Life is self-sustaining and self-generating action. Always critical parts can fail, which means life ends. You could make copies of yourself, and you could change yourself, but now, what are you sustaining? Death of what? New possibilities come with new definitions of the terms. What lives? What dies?

* If all of our material needs are eliminated or fulfilled for virtually free, what happens to the virtue of productivity?
Here is your false premise: that there will not be life forms that are as intelligent as you that make decisions that result in your destruction, whether to their benefit or not. You will still have to compete with them. Things that are currently valuable to you, things that enable you to continue living now, will become less important, other things like information, security, trust, and of course resources/control will become more important.

* How important is political activism considering Kurzweil's predictions eliminate the need for most social welfare programs?
Right, its not worth putting much time toward them compared to other things... which other things are important?

* If all knowledge can be transferred and learned with no effort, what happens to the virtue of gaining knowledge?
Information can be transferred and learned with no effort, but its not that one wants to learn tons of knowledge or tons of relationships, but one wants to learn the ones that are both consistent with Reality and worth basing decisions on that effect your life. "All Knowledge" is misleading and false, instead, replace with "communicate faster" and "process information faster".

* Would it be wrong for Objectivists to choose to live most of their lives in virtual reality? What if the virtual environment was the equivalent of Galt's Gulch?
Here is your false premise: that virtual reality is not a part of reality. "Virtual reality" is a misleading term, it should be called "simulation of Reality with another part of Reality". How much time should you spend in simulations? To what extent is it beneficial for you to do so?

* Can a non-biological entity be conscious?
What does "biological entity" mean? Do you mean life form? Of course a life form can be conscious. Can a program be a life form? Of course, I've made such life forms myself. Are they conscious? Of course they are, they just aren't very smart yet.

On nanobots: please don't forget that nano sized things cannot be independently smart. By what neural net or processor will they sense input, process information, make decisions, and control output? Where is their power source? A "nanobot" would be more like a designed cells and viruses-- ones that have very little capability like current individual cells and viruses.

Post 4

Thursday, March 23, 2006 - 5:26pm

Link

Edit

Michael:

You said:

The information still has to be processed and an attempt made to understand it, just as it is today. You can have perfect memory of something but that does not mean you understand it and integrate it into your life and actions completely.

Kurzwiel addresses this point when considering arguments people make against computers being conscious entities. He uses the Chinese Room argument as an example, which Aaron linked to in his previous post.

Do you think what you said will always be the case? To cite another example Kurzweil used in his book, do you think Deep Blue understood chess better than Gary Kasparov? Probably not in the way we think about understanding, yet it was a better chess player than he was. Why can't this example be extended to other forms of knowledge?

For example, I work as a software engineer. Suppose in the future my brain has access to an external storage device that stores information on how to write computer software for specific problems. Also suppose that my brain has access to a processor and software that can process that information and output a piece of software for any specific problem I give it. Do I still need to understand how to write software in order to produce it?

As someone who loves to learn new things, especially about writing software, it's almost depressing to think that one day we might be able to outsource most of our thinking to computers.

Post 5

Thursday, March 23, 2006 - 6:11pm

Link

Edit

Dean:

Thanks for posting your thoughts. Although I always had a pretty positive outlook on man's future, I was really blown away by some of Kurzweil's predictions. I'm still trying to digest it all and think about how my beliefs today apply to future situations I haven't even thought about. Your comments were very helpful.

Post 6

Thursday, March 23, 2006 - 6:37pm

Link

Edit

As someone who loves to learn new things, especially about writing software, it's almost depressing to think that one day we might be able to outsource most of our thinking to computers.

Are you having the external processor do all of the thinking, or are you determining what it considers good and bad? I'd less think of it as some external thing making decisions for me, and more think of it as increasing my own processing and thinking abilities. Well, I guess it depends on the communication I/O rates between the external processors and my brain. If the I/O rate was low, I'd think of it more as another thing doing my thinking. If the I/O was high, then it would be much more like an extension to my brain. Hehe : ).

Post 7

Thursday, March 23, 2006 - 6:43pm

Link

Edit

I've read two of Kurzweil's books: The Age of Spiritual Machines and this one. If never been amazed, frightened and hopeful all at once like when I've read his predictions of the future. His scenarios of the future is very reasonable; he seems to be the only thinker who dares to say where all our technological development is heading: human immortality, omniscience and omnipotence within our lifetimes.

The only issue I have with Kurzweil is how he glosses of the issue of consciousness, in humans and hypothetically in computers. He takes for granted that level of computing power = level of intelligence = level of consciousness. Conscience is a tough metaphysical nut to crack. How are we, as material beings, able to perceive our material nature? How are our thoughts transformed into actions? When the 8-ball ricochetoff the cue ball, do they become "conscience" of each other in the same way of two people bumping meeting each other?

Another issue is creativity. Computers can be programmed to receive imput. Computers can be programmed to react to input in certain ways to further certain goals. We seen examples of inductive reasoning in computers. (See e.g. 20q.net. I played it using "Objectivism", and the closest it came was "a philosophy" and "a science".) But can computers create their own "input" and "programming" in a meaningful way. Can computers write their own source code to be something completely different than they may have originally meant to be, in the same way that human can shape their philosophy and thinking?

Kurzweil's trump card in the consciousness debate is the potential to reverse engineer the human brain. He argues that if we copy the human brain, we copy consciousness. But this approach begs the question that the brain alone is the seat of human consciousness, or that it doesn't need some supernatural "soul" to jumpstart consciousness. Kurzweil alludes to this possibility himself, citing references that the human brain is actually capable of quantum computing. Commander Data and Robbie the Robot may be harder to build then we think.

However, a merger scenario seems just as likely than independent strong AI, with computers expanding human intelligence to the point where we have all the brainpower of a computer. With calculators and internet databases constantly within our reach, to some extent, we're there already.

Post 8

Friday, March 24, 2006 - 5:23am

Link

Edit

At the risk of sounding way too detached and sci-fi for a moment, I see in the next 50 years there being an inevitable race:

If BMI (brain machine interface) advances more quickly than AI, then we'll migrate to having the brain-attached math coprocessor, integrated PDA, the wireless Google coprocessor - and generally move towards a cyborg-type 'computers enhancing humanity' scenario. Enhanced humans will survive even as AI then advances as they'll be an integral part of us. However, if AI advances faster than BMI, then humans at some point will be in direct competition with independent machines more intelligent than themselves, and will (eventually) lose.

I have a slight preference for the first scenario, where silicon meets neurons. It's most inspiring to follow news such as neural interface to a joystick, a 1024 pixel camera connecting directly to optic nerves, or a Cambridge researcher experimenting directly with BMI in his own arm.

Post 9

Friday, March 24, 2006 - 5:52am

Link

Edit

Aaron:

You said:

However, if AI advances faster than BMI, then humans at some point will be in direct competition with independent machines more intelligent than themselves, and will (eventually) lose.

Would it then not be in our best interest to impede the progress of AI until we progress further along the BMI path? Not saying I think the answer is yes...just throwing it out there...

Post 10

Friday, March 24, 2006 - 9:56pm

Link

Edit

Do you think the AI side is advancing quickly enough that we have to worry about impeding it? :)

Post 11

Saturday, March 25, 2006 - 10:43pm

Link

Edit

Would it then not be in our best interest to impede the progress of AI until we progress further along the BMI path? Not saying I think the answer is yes...just throwing it out there...

I think there isn't much you can do to slow down or speed up the development of one technology verses the other. I think AI technology (boy do I hate calling it "AI" as if it were artificial) will out-pace neural-computer interface technology. But really all we need are large arrays of human bio-compatible neuro probes : ). What are the material scientists and nano-manifacturing scientists doing?

I think our best bet is to live lives of integrity so that the new life forms won't feel threatened and want to kill you but instead would prefer that you continued to live. Other options for increasing our intelligence include using designed viruses to change our genetics... which could lead to all sorts of possibilities. Surely the machines will still have some uses for some homo sapiens (at least at the start), like we have uses for other life forms.

... Its going to be an exciting future. : )

Post 12

Monday, March 27, 2006 - 1:49pm

Link

Edit

I think this is all rather overly optimistic. Has anyone produced a self-replicating machine yet? Has anyone even come remotely close? The answer is no.

Are there any conscious artificial entities yet? No - it is not clear that pure computing power will do it, though I don't discount it as possible. I just think there are too many very large IFs that make such wild predictions almost impossible. They do make for good books, though, and I prefer the optimism to the dystopian futures all the post-modernists keep predicting.