Unprecedented Suddenness

On page 6 of Engines, it is claimed that the rate of progress will suddenly accelerate:

Nevertheless, advanced molecular machines will eventually let them build nanocircuits and nanomachines as easily and directly as engineers now build microcircuits or washing machines. Then progress will become swift and dramatic.

Why? The first sentence would seem to imply that progress at the nano level will be just as swift as progress in designing microcircuits and washing machines. Unless there is something going on in the washing machine industry that I haven't heard about, this is not too encouraging. If progress in designing washing machines is a calibration of progress in designing nanosystems, then there isn't going to be an Assembler Breakthrough.

Microcircuits are a more active field than washing machines, but here too progress is more or less linear, or at least not explosively recursive. Many new circuits are designed each year, but there is no reason to anticipate that progress in designing microcircuits will suddenly become "swift and dramatic," and lead to some kind of unprecedented breakthrough. If progress in designing microcircuits is a calibration of progress in designing nanosystems, then there isn't going to be an Assembler Breakthrough.

In fact if anything calibrates nanotechnology, there isn't going to be an Assembler Breakthrough. The Breakthrough, by its very nature, has to be incommensurable with the reality of the ordinary world we live in.

The Breakthrough will happen when we have universal assemblers that can "sidestep the traditional problems of materials and fabrication," and AI systems without interfaces (i.e. they don't have to be programmed and don't have to be paid either). In other words, never.

It's not a question of what can or can't be done with atoms, or what can or can't be done with computers, or how long it will take. The idea of the Assembler Breakthrough isn't just technically wrong, it's logically incoherent. There is no such thing as a machine that can sidestep the problems of materials and fabrication, there is no such thing as a mind that doesn't have to interact with the world to ground its language, and there is no such thing as an AI system without an interface. There are no Genies.

As I have demonstrated, nanotechnology is not incommensurable with other technologies. Everything about it can be calibrated, including design -- designing nanosystems is just as hard, and takes as long, as designing anything else. The difficulty of a design problem depends on the complexity of the task, not on the size of the parts. Of course the pace will speed up, but when it does, it will accelerate continuously and stay within the same range as other technologies. There isn't going to be a discontinuity or "Singularity."

Nanotechnology will take its place among other technologies, step by step, over a period of decades. Like any other technology, it will enable us to do some things that have never been done before, and like any other technology, it will leave the economy and the world in general with the same basic structure that they have now.

On pages 49 - 50, in the section called The Assembler Breakthrough, we find this:

As nanotechnology advances, there will come a time when assemblers become an imminent prospect, backed by an earnest and well-funded development program. Their expected capabilities will have become clear...

Years of quiet design progress may well erupt into hardware with unprecedented suddenness in the wake of the assembler breakthrough.

On the contrary, as the calibration meme becomes prevalent in the nanotech community, there will come a time when the idea of a "universal assembler" and the belief system that goes with it can no longer be taken seriously. Belief systems are amazingly resistant to change, but they do change, especially among scientists and engineers who are open to the idea of a reality check.

A few years from now, the believers who are still hoping for the Breakthrough will be an increasingly isolated little sect, like the Jehovah's Witnesses who keep moving the date of Armageddon farther into the future, year after year. (If you think I am exaggerating the importance of the Breakthrough with a capital B in the nanotechnology belief system, please check out the last paragraph of the last chapter of Engines.) Meanwhile it will have become obvious to everyone else that the very idea of the Singularity is nonsense. There is no such thing as an event that is incommensurable with everything else. This may have been obvious to some people all along, but many of us could only arrive at this conclusion the hard way, by thinking the whole thing through step by step. It certainly wasn't easy for me.

The World Made New meme will give way to the industrial nanotech meme. (I am using the m-word ironically, by the way.) As the field loses its aura of utopian flakiness, it will progress faster. Scientists will be more willing to call themselves "nanotechnologists," since the word will have a different connotation. We can turn our attention to the problem of how to create industrial nanotechnology and medical nanotechnology -- i.e. how to use atomic positioners in chip factories, how to make made-to-order cells to carry out specific tasks, and how to write computer programs to assist us in designing and building such systems.

Then some of the awesome things that have been predicted for nanotechnology will indeed come to pass. For example, cell repair machines will exist. Cell repair machines don't require diamondoid replicators or AI programs writing better AI programs. Cell repair machines will be cells -- highly modified cells, yes, but still cells. Cell repair systems will be an extension and refinement of the immune system we already have. This will be the topic of another Web site.

Lyle Burkhead
March 12, 1999

 Note added March 13, 2000.

The site has been up for a year now. In round numbers, about 3,000 people visited the home page in the first year, 600 got as far as Section 7, and 300 stuck it out all the way to the end.

Here is a typical comment from a reader. This arrived just a couple of weeks ago:

Your arguments are not stronger than those of Drexler. Indeed they are somewhat weaker. And your counter-arguments are almost worthless. Well, that's what I think and I might be wrong anyway...

1) You use very simular approach to the one of Drexler. But the logic is against you. Look, when Drexler says: "we can make nanobots because they do exist - they are proteins", he is right. And when you say: "we can't make self-replicating nanobots which can produce everything because they do not exist", you are making a logical mistake. That's my strongest point.

2) Another common mistake is false analogies. Like in case of Chinese as robots. If I can replace Chinese by robots, I would remove from their brains the desire to spend money on themselves. And this will reduce consumption dramatically. Then all Chinese robots that were producting goods for INTERNAL consumption will be free to do something else. Like 2$ screwdrivers, which will be in fact FREE screwdrivers, since they are the surplus of Chinese economy.

If you don't understand the two points above or do not agree with them, then either you are an idiot (nothing personal and sorry for my rudeness) or I didn't explain them well. In latter case I will be happy to try again in some other way.

Sigh...

I have also gotten some intelligent comments, but only a few. (Now I have removed my address from this page. Comments like the one quoted above should be sent to the Extropian list.)

When this project was in the planning stage, Chris Peterson was my liaison at Foresight. We exchanged several e-mails about the project starting in the summer of 1998. I sent her a draft of Nanotechnology without Genies in December of 1998, and notified her when the site appeared on the web in March of 1999. But as far as I know, she never read it.

As far as I know, Eric Drexler never read it either. His only acknowledgement of its existence was the following paragraph, which appeared in one of the "From the desk of Dr. Eric Drexler" letters that are sent to Senior Associates (letter #25, 1st quarter 1999, page 4). This site is mentioned in passing, in a discussion of a program called Critsuite which can be used to add graffiti to web sites. I quote in full:

Will Ware and Josh Hall have made a couple of annotations on a site done by Lyle Burkhead in which Lyle makes skeptical points about nanotechnology. To join in, see http://crit.org/http://www.geniebusters.org

You can use that URL to see the level on which the subject is discussed at the Foresight Institute, when they discuss it at all.

Naturally, there is no link from www.foresight.org to www.geniebusters.org.

Several people have said that I am attacking a position that no one believes anyway. Engines of Creation has, supposedly, been superseded by Nanosystems. This is not true. Eric Drexler has never backed off an inch from his original vision. To verify this, see his Introduction to the Web version of Engines of Creation, where he says

There's not much I'd change in Engines if rewritten today...

I know from the mail I get, and from discussions in other places, that there are still people who believe in the scenario described in Engines of Creation. There is a steady stream of new converts, most of them very young. This will probably continue for several more years.

June 27, 2000

Meanwhile, however, on the memetic front, the battle is over, and it turned out exactly as I predicted when I first wrote "Genies." Earlier on this page I said that there will come a time when the idea of a "universal assembler" and the belief system that goes with it can no longer be taken seriously. That time has already come. In the June 13, 2000 issue of Nature, page 730, there is a news feature about nanotechnology, which starts out as follows:

There is no such thing as bad publicity, it is sometimes said. But ever since Eric Drexler brought the term 'nanotechnology' into vogue in his 1986 book Engines of Creation, some researchers have felt that the field has been burdened by unwanted baggage. Drexler envisioned an era in which factory production lines were replaced by self-replicating, nanoscale 'assemblers' -- and warned that such entities could supplant humans to become the dominant 'life' forms on our planet. These ideas were quickly seized on by transhumanists -- people who imagine what the world will look like after technology has rendered us extinct.

But today nanotechnology is acquiring the respect researchers in the field believe it deserves...

There is no way to recover from this. The only way Eric could vindicate himself would be to go ahead and build a universal assembler, and that's not going to happen.

---

I have had some second thoughts about all this, which can be found on the Afterword page and the Singularity page. There isn't going to be a Singularity as originally conceived by Eric Drexler, Vernor Vinge, and others, but that doesn't mean TEOTWAWKI isn't going to happen at all.

back to the table of contents

home

---