Welcome to the Soft Machines weblog, which I hope will be the more interactive part of the overall Soft Machines website. One purpose of the website is to publicise my book, of course, but I would also like to make it a useful resource on nanotechnology. I’m not going to try to be comprehensive – there are many other good resources on the internet and elsewhere. Instead I’ll give a personal view, from the perspective of a scientist working in the area, of what I think is really interesting. I’ll also use the web-site if I need to amplify points in the book, to report feedback to the book and my response to it, and if necessary to update and correct it. Early drafts of this material will probably make their first appearance on this weblog.
4 thoughts on “Soft Machines weblog”
Comments are closed.
I thought I’d contribute my first feedback on your book here, since I couldn’t find anywhere specifically for the purpose on your website. Please move them if there’s a better place.
I just started reading your book and decided to comment as I go rather than save it all up for the end. I’m very interested by it; a book by a practising scientist who’s prepared to put forward his own opinions on a subject where so many are keeping their heads below the parapet.
I’m presently reading the introductory material; I find it’s often a good way to ‘calibrate’ a book. So far it’s generally a good survey of several disparate areas.
I’m a computer software engineer, and there were two points in the computer section that I felt obliged to comment on.
Firstly you say on p41 that it is necessary to have an inverter and a NOR gate to build a computer. If you consider the transfer function of a single-input NOR gate, you’ll see that it is an inverter. Only one gate type is necessary. The important point is that as well as logic, it is necessary to provide amplification to overcome losses. Many proposed novel techniques are claimed to exhibit a useful transfer function, but have a lower output energy level than the input. Large-scale practical circuits cannot be built on this basis.
Secondly, you say on p50 that computers do not get much cheaper. I don’t think you can have done much investigation before making this statement and certainly you have ignored Bell’s Law! That’s even less of a law than Moore’s Law but Gordon Bell’s marketing insight that each model of computer should be replaced by two in the next generation – one with the same performance for less price and one with the same price but higher performance – has strongly influenced how computers are sold and perceived.
Then I started to read about fluid dynamics. I’m a sailor, so I have some interest in this subject.
I was tired when I read p57 and got very confused by the discussion of area and a symbol ‘a’. It took me a long time to puzzle out that ‘a’ was not the area but the length!
More interesting was the discussion of flight and its scaling properties. Two thoughts occurred to me as I read it.
Firstly, many insects don’t ‘fly’ in the sense you used it. Some stay aloft entirely using drag forces, rotating their wings to change the effective area on the up- and down-strokes. An expert lecturer at a Royal Aeronautical Society meeting said there are as many different mechanisms for flight in insects as there are species of insect.
The second thought was that you completely ignored buoyancy. Balloons can stay afloat without expending energy. What’s the scaling law for a balloon?
My point is that it is difficult to claim something is impossible. You do have to consider all possibilities (and prove that you have done so 🙂 There are many aspects of ‘Prey’ that I find totally unbelievable, but the possibility of small objects staying aloft is not one of them.
I look forward to reading the rest of the book.
Regards, Dave
Thanks for your comments – I’ll reply to them here for the time being, then if I accummulate some more comments I may move them to the main website.
Thanks for the comments on computing, and particularly for pointing out the importance of amplification. Maybe I should have made more of that. As for the extent of research on PC pricing, I’m guilty as charged – my research consisted of remembering roughly how much each PC I’ve bought at home or work has cost, from the 80286 clone I used as a graduate student to run a multichannel analyser, to the G5 I’m writing this on. But I still stand by the broad outline of what i said; computing power has increased by orders of magnitude, but the unit cost of PCs has not decreased by orders of magnitude.
Sorry about the a in Reynolds number. It’s the symbol often used in the technical literature but you’re quite right, it’s confusing; I’ll replace it with something else if I get the opportunity.
As to flying, the key point isn’t whether small objects can stay aloft; they clearly can, even without buoyancy, thanks to the importance of drag. What’s interesting is whether they can propel themselves forward – remember, in Prey, the cloud of nanobots didn’t just drift about in the wind, it was able to chase after things. It is this flying speed that has such an unfavourable scaling with size. You’re right to say insect flight is very complex – the Stephen Vogel book “Life in Moving Fluids” is a good introduction. But that doesn’t mean that one can’t make some robust conclusions.
I’m sorry to take so long replying. I don’t have time to browse interminable websites, nor to subscribe to unlimited RSS feeds, so I had expected the courtesy of an email alert at least. You can reach me at either Dave.Howorth@acm.org or dhoworth@mrc-lmb.cam.ac.uk.
With regard to computer pricing, I think you’re simply ignoring the evidence. Here’s the first data on perfomance vs time I came across with Google tonight (totally unscientific but at least impartial and impersonal). Where’s your support?: http://www.neurophys.wisc.edu/ravi/docs/m122p1.pdf
With regard to flying, I think your answer creates points not present in your book. The book discusses how small objects can stay aloft using lift, specifically stating that ‘in addition to finding the power need to move forward, a flier needs to generate enough lift to stay in the air’. I dispute staying in the air is a problem, and you simply reverse your argument! Which way do you want it?
I’m probably as sceptical as you regarding the nanobots in ‘Prey’ but in terms of forward motion, the obvious answer is to increase the Reynolds number. In other words, they must link together such that their effective size is greater. Such a strategy is not precluded by the book, IMHO. (even though the whole concept of the nanoswarm is a non-starter)
Cheers, Dave
PS Why could I not find my comments by searching your site, but only by googling the web? Are you censoring comments?
Dave, the graph you link to shows (as far as I can tell) the falling cost of computing, not the falling cost of computers. Since the introduction of the IBM PC/XT, the cost of computing has fallen by six orders of magnitude (if we allow ourselves a bit of extrapolation). Guessing that the PC/XT cost £2000 or so, does this mean that we can buy an PC/XT now for 0.2 pence? No, of course not, and this is my point.
I’m sorry if what I wrote in the book about insect flight wasn’t clear. I don’t actually have a copy at home so I can’t check exactly how I phrased it. If you are interested, once I again I recommend reading the Vogel book. I don’t think your strategy of the flying nanobots liking together will be helpful. Certainly, if you create a swarm of tiny things, each of which is separated by only a small multiple of their radius, then the fluid between them will effectively be entrained by the swarm, which will indeed behave in its fluid mechanics like a larger, solid object. But by the same token, this means that the objects in the centre of the swarm have no way of contributing to its overall motion, and you will have to rely for propulsion on the small fraction of objects at the swarm’s surface.
I do censor comments, but only when they are selling on-line poker or herbal breast enhancements. Your comment, and my reply, was exactly where you left it. As you discovered, the search function only works for posts, not comments. That’s how the software comes; I have made changes to the way the site works in response to requests from readers (for example the latest post feature) but I haven’t got the time or expertise to make extensive tweaks. I’m very happy to correspond by email, but if you start an exchange in this public forum I’m going to assume that’s where you want to continue it.