Nanotechnology without Genies

The Geniebusters argument is a dialectical argument. The pages are not independent of each other. There is a layer effect, kind of like a 3D stereo picture. If you read the pages sequentially and then look over what you have read, you will (I hope) see patterns that are not apparent at first glance. This applies particularly to Parts Two and Three.

1. Introduction (on this page, below the table of contents)

Part One -- The Truly General Fabrication System

2. First example of mapping and calibration: the Shapeshifter
3. Calibrating the Universal Assembler - the programmable organism
4. Calibrating the Universal Assembler - the computer as genie machine
5. Calibrating the Universal Assembler - the economy as genie machine
6. The devil is in the interface
7. Nailing down a Conclusion about Universal Assemblers

Part Two -- Either program them or pay them

8. Fine-grained points in a bulk economy
9. Atomic positioners in manufacturing
10. Replicators are not free
11. Automation is not free
12. The "box" is a society
13. Subsystems in the box are corporations
14. Genies, large and small

Part Three -- On the Internet, nobody knows you're an AI

15. No Moore's Law for software
16. The irrelevance of nanobots
17. The irrelevance of robots
18. Programs writing programs
19. Robotic Microserfs
20. The bootstrap problem
21. The Irrelevance of AI -- the Reverse Turing Test

Conclusion of Part Three: The Singularity (This wasn't part of the original article, but the Singularity page is a continuation of the line of thought in Part Three, so it belongs here.)

Part Four -- miscellaneous

All machines are specialized

Diamond as universal material

Positional synthesis vs. chemistry

Autarky is unstable

Conclusion

Afterword
With some second thoughts, updated June 29, 2000. This was the turning point. This was when I decided to abandon my "nanocritic" role.

Coherent energy from the quantum domain - a new page added in 2004

A sarcastic look at universal assemblers - another new page. This page also has some comments about investing in nanotechnology.

A Genie Story (on a lighter note)

I get a report from my ISP which tells me, among other things, the search terms people use to arrive at this site. I get more inquiries about genies than about nanotechnology. Some of them don't even know how to spell it. My favorite search term is "genis book of world records."

And now, here is the geniebusters article:

Introduction

This article is a critique of Eric Drexler's ideas about nanotechnology, but not the same kind of critique that has appeared in the past -- in the Scientific American discussion, for example.

A lot of people read Engines of Creation and think: There has got to be something wrong with this. But they can't put their finger on it.

They always assume that if nanotechnology is possible at all, then everything in Engines of Creation follows, and so they think they have to show that it is impossible to build machines out of atoms. These nanocritics lose the argument every time, because in fact it is possible to make machines out of atoms. The problem lies elsewhere.

I take a different approach. I acknowledge the obvious fact that nanotechnology will exist. It is already well underway. It seems like almost every issue of Nature has an article about nanotechnology (in a general sense). However, the fact that nanotechnology will exist does not imply that little robots will supply all our needs for free. In other words, Nanosystems may be true, but this implies nothing about Engines of Creation.

I have gotten quite a bit of feedback from various people, many of whom have somehow gotten the impression that I am attacking the very idea of nanotechnology. So, at the risk of being repetitive, let me spell it out as clearly as I can: this article is not a critique of nanotechnology as technology. It is a critique of the belief system (the "nanotechnology meme") that has come to be associated with it.

The disagreement is not about whether nanotechnology will exist, but about what it will look like and what effects it will have. It may help to consider an analogy -- Newton's Third Law of mechanics. This law doesn't put any limit on how much force can be exerted. Newton didn't say anything about what can or cannot be done. He simply asserted that there is always an equal and opposite force. That's the kind of point I am making here. I'm not saying anything about what can or can't be done, I'm just making a general point about the structure of the situation.

The position I am arguing against was summed up at the end of the first chapter of Engines of Creation, in the section titled The World Made New, page 20 --

To have any hope of understanding our future, we must understand the consequences of assemblers, disassemblers, and nanocomputers. They promise to bring changes as profound as the industrial revolution, antibiotics, and nuclear weapons all rolled up in one massive breakthrough.

The consequences of this breakthrough are described throughout the book, for example on page 63 --

In short, replicating assemblers will copy themselves by the ton, then make other products such as computers, rocket engines, chairs, and so forth... Assemblers will be able to make virtually anything from common materials without labor...

and on page 81 --

This transformation is a dizzying prospect. Beyond it, if we survive, lies a world with replicating assemblers, able to make whatever they are told to make, without need for human labor.

and on page 94 --

Further, with robotic devices of various sizes to assemble parts into larger systems, the entire manufacturing process from assembling molecules to assembling skyscrapers could be free of labor costs.

These assertions are every bit as absurd as they appear to be, but not for the reason most nanocritics would give -- i.e. not because it's impossible to make machines out of atoms. Regardless of what can or can't be done with atoms, the assertions quoted above are still absurd.

My position on nanotechnology is:

1. Nanotechnology is commensurable with other technologies, and it will progress at the same rate as technology in general. Industrial nanotechnology will emerge gradually and continuously from the technology of the present. There isn't going to be any such thing as a microscopic, self-contained "universal assembler."

2. Automated systems always exist in a larger context which is not automated, and their products are not free. Regardless of what can or can't be done with atoms, or with computers, machines will still have to be designed. They won't design themselves. Designing, building, and retooling complex apparatus won't be easy or free in the future any more than it is now.

3. Artificial intelligence is irrelevant. Whatever capability AI has at any given time, humans assisted by computers will have already reached that point and moved ahead.

4. Making things with atomic positioners will be at least as expensive as making them with biotechnology or bulk technology. There isn't going to be any such thing as "molecular manufacturing," if that expression is construed to mean using atomic positioners to make common objects and materials that can already be made in other ways. Atomic positioners will only be used to make things that could not be made in any other way.

In other words, the nanotechnological TEOTWAWKI isn't going to happen. There isn't going to be a nano-utopia.

Without AI systems ("Genies") that design everything for free, molecular manufacturing is just agribusiness.

Nanotechnology will indeed have profound effects, but not the effects predicted in Engines of Creation (except the effects described in chapter 7, which I agree with, up to a point). With nanotechnology we can build apparatus that will enable us to understand the phenomenon of quantum entanglement, and make use of it. This is the key to the door that leads to wonderland. (This "apparatus" could exist within a cell. All kinds of nanodevices could be built within cells. In fact that's going to be the easiest way to build them.) Mesoscopic physics holds many surprises. Biology has been the happening science in the last several years, but physics is going to reclaim its place in the sun.

Transhumanists have to consider which research agenda we should be pursuing. Should we be working on universal assemblers, or made-to-order cells? Should we pin our hopes on computers "waking up," or should we concentrate on perfecting our own brains? Should we spend our time designing nanoscale gears and bearings, or studying how intracellular signals are regulated?

I think Eric Drexler's molecular machines are horseless carriages. It's not that they can't be built, it's just that they will turn out to be irrelevant.

The nanotech meme has (at least) two parts that can be considered separately:

1) the idea that molecular manufacturing plus AI will lead to a cornucopia,
2) the idea that our cells can be perfected.

The first idea is an illusion, but I agree with the second idea, and obviously it has profound implications. The world will indeed be made new. When our cells (including our brain cells) are perfected, the result will be the "superhuman intelligence" that the Singularitarians are looking for. It isn't going to emerge from an AI lab. The recursive center from which ultraintelligence is emerging lies within us.

I want to say just a little about the form of the argument.

In mathematics, one frequently uses inequalities. I may not know much about a certain function, but if I can prove that it grows faster than a linear function but more slowly than an exponential function, that's a step towards getting a handle on it. Mathematicians also map one structure onto another, perhaps from a different area of mathematics, and the mapping illuminates both areas. The simplest example of this is analytic geometry, where we have both algebraic and geometric structures, and the interplay between them helps us to understand both algebra and geometry.

In the following discussion I am going to be using comparisons, similar to mathematical inequalities (e.g. arguments of the form "this must be at least as complex as that"), and mappings from one domain to another. I am also going to invoke principles such as

everything exists in an economic context
all machines have interfaces
complexity is independent of size.

Eventually I intend to have a much longer discussion of this, since the principles involved are more important than the specific argument about nanotechnology. This subject deserves a more systematic treatment than I have given it here. Time will settle the argument about what nanotechnology is going to look like, but meanwhile the fallacious thinking that led to Engines of Creation will continue to mislead us in other ways.

Ultraintelligence will be possible only if the mind has a correct model of itself. Engines of Creation is plausible to a lot of people (including me, when I first read it) because we have a fundamentally false model of ourselves.

For anyone who aspires to ultraintelligence, philosophy is not a side issue. Logic, semantics, and philosophy of mind are just as important as mathematics, biology, and computer science. All these subjects should be combined into one unified curriculum. That isn't going to happen in universities, but there is nothing stopping us from making that unification in our own minds.

And now, let's proceed to the argument. I am going to assume that the reader is familiar with Engines of Creation. At the very least, you should have read the section titled Universal Assemblers on page 14. The whole book is available online - naturally, just like the Bible is available online - so if you haven't already read it, you can follow along.

I begin with three riffs on the theme of "an entity that can make anything, including copies of itself."