Nick Szabo  🔑

After all this time Drexler & his myriad fans have still failed to convince the vast majority of chemists & physicists & more importantly failed to convince reality in the form of actually working machines, that he has figured out a way to cheat Heisenberg on such small scales.

What do you think the uncertainty of position of a carbon or even hydrogen atom is? Have you bothered calculating it? Hint: it isn't big. Just do the math if you don't believe me. And furthermore, you can IMAGE them.

Curious how you do such a calculation? Standard quantum limits (e.g. in interferometry) usually involve specifying a free variable, typically some type of frequency. Are you calculating a standard quantum limit in your calculation?

As far as I'm aware, it's standard physics that both position and momentum (but ofc not both at once) variance can be reduced arbitrarily. Indeed, LIGO has begun using squeezed states (of light, not matter, but the idea is the same) to improve their strain sensitivity.

Mechanical devices have to be simultaneously precise as to both position and momentum.

That puts severe limits on how small they can be made and retain sufficient precision.

So? We're not talking about electrons here, we're talking about things with many orders of magnitude more mass. There are actual calculations in Chapter 5 of Nanosystems. Which of them do you think is incorrect?

It is not enough to say "quantum mechanics!", there are actual well understood equations that govern quantum systems, and you can do calculations using them, and the calculations say you're wrong. If you disagree, say which eq is wrong in Ch. 5 of Nanosystems.

Those are the concepts we presently use to describe them for certain purposes, yes. To more clearly conceptualize what I'm trying to get at, consider the difference between Drexler's "molecular machines" and the molecular dynamics selected by evolution...

If complex molecular assembly were not possible, we would not exist. But this is why I say the industrial understanding of crystalline / rationalized systems is the wrong way to think about molecular assembly. Rather, we require a kind of system's intelligence capable of...

Exploring possibility space in a manner that balances "blind" evolution with the intentionality and time scales of industrial production. I do believe it's possible to make significant progress as constrained by the tension between these two approaches.

Which is why I'm hopeful for the future of projects such as AlphaFold, given that this is precisely the kind of constraint problem it's designed to approach.

Perhaps a good summary is that nanotechnology as originally proposed is not feasible, but synthetic biology aimed at similar ends may well be.

Biology achieves very different ends by very different means.

You can easily do the calculation and find that the uncertainty principle is totally unimportant in this domain. You can even read the calculation in Nanosystems. If you're going to make a factual claim like this, then you should at least know the literature.