While you work on these problems I’ll be over here worried about the pollution-like side effects of ordinary endothermic nanomachines.
They're probably in local maxima, not in any global maximum, but it's also likely that there is no single global maximum given how different the requirements are for using different energy sources.
The weak bonds is what complex chemistry and thus complex functionality possible. Stronger bonds is dead matter, and dead matter does not compete.
There are a few fluorinated natural products made by wild organisms, but they are very rare considering the elemental abundance of fluorine on Earth and the high utility found for fluorine in pharmaceutical development:
https://www.ncbi.nlm.nih.gov/pubmed/23034231
The strong bonds formed by fluorine, and its formation of insoluble compounds with alkaline earth metals, probably explains the rarity of natural fluorinated products. The fluorinated natural products are no more "dead matter" than other poisons evolved by plants against herbivores, but it's hard to evolve out of local minima that exclude fluorinated compounds.
We are art made by microbes. (I like to think.)
ie. It's not an argument against grey goo, it's an argument the grey goo catastrophe rules everywhere and has done so for eons.
This is like saying "no bird can fly faster than the speed of sound therefore flying faster than the speed of sound is impossible."
Grey goo, on the other hand, has to be self-sufficient and sustainable to exist.
I wouldn’t call that “grey goo”, but the threat model is a spectrum not a Boolean.
Without any immune system you’d rot like a corpse.
Anything that can metabolize lots of kinds of multicellular living matter could can do insane amounts of damage even if it's not a pathogen... for example agriculture is quite frail and we depend on it!
It’s right there in the first sentence. Nothing about eating concrete or sand.
That is, the tinier a machine, the more susceptible it is to breaking down due to wear and tear. And the more energetic a machine, ditto. So incredibly tiny, incredibly energetic machines, would have a minuscule lifespan.
And that's not even touching on the issues of power supply and storage.
I think AI is a much more existential threat than gray goo.
In reality there are limits; there's no reason to think technology can improve indefinitely. We may still be far from those limits, as e.g. Drexler's books show, but fundamental rules such as "Effort requires energy input" are unlikely to ever change, and it's equally unlikely that most modern scientists would regard limit-tech as being magic.
As for the average guy on the street, he probably considers computers to be magic...
Even extrapolating Moore’s Law’s exponential growth of x2 per 1.5 years would still take ~200 more years to reach the limits of computation — Improvements to particle accelerators aren’t anywhere near as rapid — so we have a long way to go before we reach the end of experiment even if we improve as fast as possible and have almost no surprises.
Not that I disagree with your general premise that some massive surprise could be waiting for us; I happen to think there is, and I get the impression that it’s totally unconventional in the physics community to say “at least one of GR and QM is wrong, but we don’t have any experiment that tells us which”.
I hazard that should we find such a surprise, it will be under somewhat extraordinary conditions not commonly found in nature, and quite likely ill-suited for power generation on the nanometer scale.
I’m not a betting type, but if I had to take a bet I would expect it to be ~”and this is how the conditions of the Big Bang can be recreated and exploited to make something that looks like a star but which will exist forever. Also the interior of our fake star is now a new universe 50 billion light years across even though the outside circumference is ten light seconds.”
Not very useful for cell-scale engineering. Might count as “magic”, but the bigger-inside thing is a serious idea for making the Alcubierre drive less unphysical.
Our level of understanding seems, broadly, a better level of understanding than a 16th century layperson would have of topics like the existence of electromagnetic waves and electricity, let alone the batteries, semiconductors, and software applications of semiconductors which ties them together.
That is, very few people know -- or care to know -- how their cell phone works. Those on this forum, of course, do. But, to the average man-on-the-street, its magic.
The difference is that a modern person knows that if they really wanted or needed to, they could find resources to explain how their phone works (beginning with electricity, electronics, programming, etc.)
The 16th century guy would be at a disadvantge for he'd correctly intuit pretty fast that nobody on the planet knows how it works and there is nowhere to get the info.
Any sufficiently advanced technology does not break the laws of physics anyway.