• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed

  • Powered by FeedBlitz

« Molecular Gearheads | Main | Nano Culture Clash »

March 16, 2004


Feed You can follow this conversation by subscribing to the comment feed for this post.


yes, but, you still need a.i. advances to really make nanotech move.

I do believe that the general idea is that mnt is used to find the 'real' a.i. by just reverse engineering; that is still going to take awhile after whenever somebody makes something with a bunch of nanoassemblers.

Not only that, but, while mnt is developing before the true a.i. comes to put nanotech in really high gear, everybody will be right around each other as far as mnt is concerned; i mean all the countries of the world who are bound to work towards mnt.

I think it is still going to be awhile before any supercharged mnt leaves everyone else in the dust.

Robin Green

Please explain why you think AI is needed to do nano, David.


What I am meaning here isn't just that quantum chemistry software is needed for mnt, but to do the type of mnt that Eric Drexler talks about in his Engines book where innovation and discovery come at such a speed that only superadvanced mnt A.i. can understand; i'm talking about the spike and singularity point people like John Smart talk about.

I would also point out that this type of a.i. is going to be whats needed to really do the medical benefits of mnt; your going to need that advanced mnt-a.i. to constantly monitor, analyse, and fix an active lifeform, especially one as advanced as a human being.

Chris Phoenix, CRN

AI would certainly accelerate things, but MNT will have huge impacts without it. Including in medicine.

Simply being able to do a complete blood workup every minute would allow medical conditions to be caught much earlier. It doesn't take AI to interpret today's medical tests. And very simple pattern-learning software, connected to rich sensors, could undoubtedly tell us a lot we don't currently know about health.

Nanotech product design doesn't have to be hard. Start with a few basic reusable building blocks: motors, bearings, snap-fit parts, etc. From there, you can build 'most anything you want to design, using software engineering-type levels of abstraction. Today's products are very clunky; it wouldn't be hard to equal them with nanofactory-built products. No AI needed for a manufacturing revolution.



Actually, the reason why I said mnt is going to need a.i. is because I see that with a.i., we can get the same ability to make things without hugh amount of human labor by applying sufficiently sophisticated a.i. to even microtechnology. I've heard of the use of lasers to manipulate the placement of atoms - microfabrication using a kind of micro-water jets; i've read they can get pretty small sizes with this - and then there is another use of lasers that can be used to effect the shaping of materials; i forget its name, but i've also read they've gotten to pretty small sizes. Applying sufficiently advanced a.i. to any of those manufacturing ideas can give much the same effect as a black box mnt or even a freely moving around assembler.

Well, you see, I believe I have some a.i. ideas! I know this sounds rather brash. My idea is kind of like this - i see a gene as a kind of an idea, and the bacterial swapping of genes as the aquiring and mixing up of ideas. I actually have some other what I would call insights on intelligence which fits the above idea and makes me think I'm on the right track; yes, I know; this is not experimentally and peer reviewed confirmed, but I think I've given enough indication that I do have some ideas towards a.i. I only wish I knew my programming better which I am working on.

If I were to develop my ideas and then use them to guide some of those manufacturing ideas above, I could make myself a rocket without mnt without asking anybody for money and all the other benefits of mnt that Drexler points out in Engines and even Unbounding.

Chris Phoenix, CRN

David, it sounds like you're reinventing genetic algorithms. (Do a google search.)

I agree AI would help to achieve some advanced manufacturing without MNT. But that's a separate question from whether MNT will need AI.

The comments to this entry are closed.