Is this possible in the future?

Ryan_m_b

Interesting stuff! Do you have any plans to take this to a more complex level? I.e. a hand that replicates the function of a biological hand in terms of freedom of movement, digits etc.

Hells

What kind of movement can you identify?

hisham.i

No plans yet:D
this is the link hells you can see the video
i classified
grip close and grip open
wrist flexsion and wrist extension

http://www.youtube.com/watch?v=QOfSV...eature=feedlik

SW VandeCarr

Here's a critique of "Kurzweilian" prophecy by a leading neuroscientist (on Kurzweil's own blog). The point is that Kurzweil's time lines are, to say the least, very controversial and do reveal a lack of understanding of neuroscience. Kurzweil makes no secret of his hunger for advanced biotechnology. He wants to live forever with a supercharged brain.

Having said that, if we ignore his time lines, to argue against his basic thesis is to argue that there are prescribed limits to technological advances and that certain things can never be done. Provided these things are in principle possible and do not invoke magic or mysticism, there is no scientific basis for challenging his notion of a "singularity". We are living in a "post singularity" age now relative to what Isaac Newton could have predicted in his most speculative moments.

http://www.kurzweilai.net/the-singul...m_medium=email

Ryan_m_b

I would slightly disagree with this, the notion of the technological singularity makes the implicit statement that the future is unknowable due to the creation of superhuman intelligences. Whilst this is analogous to the past I don't think this is the way that people like Kurzweil usually use the term. In the Kurzweilian (and others) use of the term it means the creation of above human intelligences that then go on to create above-above human intelligences that then go on to create above-above-above human intelligences etc etc leaving the rest of life as far away from the top of the food chain as plankton.

The singularity ala Victor Vinge relies on the creation of these superintelligences through either strong AI or upload and emulation of the human mind, I would contend that whilst these may be physically possible this is a world away for suggesting that it can happen.

For example if we take mind uploading; Kurzweils arguments are based on the idea that the human mind can be simulated without simulating a human brain and then on top of that can be safely modified to boost intelligence both quantitatively and qualitatively. I would suggest that this is currently not supported by contemporary science, the idea of simulating a human mind without simulating a human body plus environment presupposes a solution to the hard problem of consciousness and further presupposes that this solution allows for modification.

So taking mind uploading as not being a viable thing to propose (whilst it is supported by evidence that human minds can be created it is not supported that they can be decoupled from the human brain and modified) we would have to create superintelligences from scratch. The biggest problem I see to this is ethics. Creating an AI fit for a job is like selectively breeding an autistic slave, any human equivalent AI (whatever that means) is going to need the same rights as human beings. This issue is compounded further when you examine some of the methods proposed for this superintelligence generation; e.g. genetic algorithms being akin to genocide/crimes against humanity. The ethical arguments against the generation of AI for labour are huge. Often a counter argument is that AI can be made to be as capable as a human but without emotions, ego and consciousness. This is another claim that I would say is not supported by evidence (it again runs into fundamental unknowns about philosophy of mind, mainly are human capable zombies possible).

So whilst the idea of a singularity doesn't break any current scientific consensus the same way perpetual motion, cold fusion and homeopathy do it still relies on a number of assumptions on subjects for which the answers are currently unknown;

If we can build strong AI
If we can upload the human mind
If we can modify human-level intelligence to something better
etc etc

SW VandeCarr

I read his book when it first came out (The Singularity is Near). That doesn't mean I'm a "convert". I think his time lines are absurd. But his basic thesis is rooted in the idea that the "hard problem of consciousness" must in principle be solvable in terms of physical processes which can be quantified and reduced to code. As code, it can be considered information which can be inputted, stored, and retrieved from a "suitable" medium. This is certainly consistent with what scientists profess to believe and probably should believe about nature in general. Of course, as you say, it may well not be problem that can solved in the foreseeable future. Maybe by 2300, 2500, 3000 or 10000 CE. Who knows? These are incredibly short time spans in terms of the history of life on earth and not all that long in terms of the existence of modern humans.

Evo

I'm sorry, but discussion of Ray Kurzweil goes against our policy of quoting or linking to crackpot sources.

You might want to read this.

http://scienceblogs.com/pharyngula/2...itarianism.php

Ryan_m_b

Sorry Evo I just want to quickly get this out then I'll shut up
I agree but there's a second and third premise after this necessary for Kurzweils singularity. Firstly he proposes that we can simulate a human, we both agree that that is technically fine. But after that he proposes that the simulation is of a human mind without a human body, this I contend since the human mind is an emergent property of the human brain which requires a body which requires an environment. Thirdly his whole idea relies on the fact that this decoupled mind can be modified to be >human intelligence and that it will then go on to modify itself to >>human intelligence etc.

So whilst I agree that it is technically possible to envision scanning and simulating a human plus environment I disagree that it is currently supported that we can modify this process to the extend required for Kurzweils singularity.