2003-04-25 Eric Drexler http://conferences.oreillynet.com/cs/et2003/view/e_sess/3804 50th anniversary *today*: the announcement in Nature by Watson & Crick of the "chief digital information storage system on the planet". 2 bits per base pair, each cell contains 1Gb of data (with exception of red blood cells). He asks: - what is the main digital storage system on your computer? -> the main is in bacteria on your computer 1Mb in 1 bacteria, and a million bacteria is very small. It's free. We don't notice it, we don't know how to make use of it. He proposed we use the same principles in technology. More electronics made in the fields of Iowa! Sunlight to electrical energy. What is shows us: This is possible. Biology exists. So some perspectives on chemistry, biology, mechanical engineering, computer science from the pov of nanotechnology. They can do stuff with molecular precision, and cleanly. He's saying: taking these principles, we should be able to do this ourselves. Strong, tough materials, etc [hang on. The fact that there's loads of data: doesn't that also mean that nanotechnology is *really* hard to program?] a bacterium is a 1000nm x 1000nm x 1000nm computer science perspective... [I missed the others] he says that computer world tells us that working with the bits rights at the bottom (bits, atoms) lead to highly reliable general purpose systems that can iterate and create other systems. digital: bonded *or* nonbonded states. [hm. isn't there loads of redundancy, to make sure things don't go wrong? componsate for cosmic ray interference etc] powerful technologies have downsides: grey goo. if we build replicating things like bacteria that don't require food, just more material, it could turn the biosphere into dust into not very long. analogy: making cars vs a car that is able to forage in the wild, make more cars, and live on treesap. so it's not really a problem. we can supress things. although military research is a concern. [so this is weird. why is he talking about all of this stuff? he's extrapolated a huge future for nanotech, and is arguing aspects of that. but the low-level stuff could perturb the levels about so much, surely it's not worth considering? my opinion. things are never that different from moment to moment. making nanotech will continue being hard and expensive, like current tech. sure it'll be cool. but if nanotech is really powerful, then it'll turn out to be so hard to program it takes like 10 years to do it. i'm a big skeptic today.] [hang on, i missed a bit] micron-size 1Ghz mechanical computer like Babbage?? Then you can fit a billion of them in a computer, and they only output 1nw each in heat. a billionfold increase [give or take orders of magnitude], there's a lot that can be done with that. he says: use it in monitors. the monitor would measure where your eyes, and direct different pictures towards different people. * CURRENT STATE of nanotech he put forward the term in the 1980s. it's pretty broad though. analogy with spaceflight: we'd be at the point where people understood you could have multistage rockets, but before you have satellites. so everyone was claiming *everything* was space technologies. towers, gliders, etc. [this is rambling. i'm going to have to get somebody else's notes on this. he's got an interesting perspective on this -- it's going to take a lot of interested people to translate his core concept (which he obviously has, and i don't doubt that it'll happen!) into a usefully memetic thing.] revolutions: - universal inking machine - universal music playing machine (CD replaces having to own a flute and a flute player) he's now mentioning a generic machine that can produce any 3d pattern, ie: a universal *thing* builder. which is a revolution, because it can produce *itself*. [so that's fine, we talk about that. but now he's talking about the problem of how to stop people producing bad things! digital signatures, DRM, etc. it seems a bizarre problem to think of this early.] Q - about DRM A - how to maximise liberty and flexibility but keeping the control Q - two questions. what's the current blockage, and will it be easier or harder to defeat DRM. [bloody hell. enough with the DRM already!] A - we haven't coupled design technology AND fabrication technology at the molecular level. but the main reason is cultural. the people exposed to the current tools we need are chemists; the mentality to build these systems are systems engineers. a large collaboration in the chemistry world is two groups -- in the engineering world, big is the Apollo program. also scientists will have to be working for engineers, but that's the reverse of currently. on the subject of control: we'll see "the emerging transparent society that we're increasingly embedded in" [embedded again!], and an understanding of what the dangers can be. Q - what have you changed your mind about in the last 20 years? A - the courage, honesty and breadth of understanding of the scientific community as grown up under federal funding. his estimation of those virtues have gone down. he's waiting for this "metastable" "free-floating set of assumptions" to "collapse" [nice words!], that's the incorrect assumptions that drexler finds "disconcerting" they're still around and blocking funding. "solar power by repaving roads" -- this sort of this is small science. small amounts of materials, low budgets. it's not an apollo programme, much smaller than that. great question from Cory: "the consumptive act is also providing, if i "take" a kitchen chair, i'm copying, so the world's support of kitchen chairs increases the future of nonrivalrous goods is there anything like an economy in this future" recommended by drexler: crnano.org another question. at the nanoscale, noise (eg radiation) can cause mutations in the machine or the information store (like DNA). couldn't this be bad? evolved systems, biological system, were designed to be evolvable. but built systems are designed to be build in big jumps, not small jumps. you do stuff like many each bit depend on all of the other bits: if any of the bits change, the decoding of the entire thing stops. but it's the kind of safeguard that can be built in. q - what's the best example of the most progress been made? a - the next generation of chips, but that's not using the molecular machine definition.