Nanotechnology and Mutation

In another online venue, I was asked this question recently: Could radiation damage bring about random mutation of nanobots?

The answer I gave was: Yes and no.

Yes, radiation damage will occur. Is that a problem? Yes, it's a problem for the individual nanobot or nanosystem that is damaged or disabled. However, current designs for molecular manufacturing do not include free-roaming nanites; that has been shown to be less efficient than a model in which all the nanobots are fastened down inside a nanofactory.

It has also been shown that vast amounts of redundancy can be built in at very low cost, making radiation damage essentially irrelevant, at least during the working life of each nanofactory (which likely will be made obselete by a more advanced model within months).

As for random mutation, remember that each case of radiation damage is far more likely to be disadvantageous rather than advantageous. So, right from the start, the odds are stacked against progression via mutation.

The other important point is that evolutionary mutation only applies in cases of reproduction, as in biology with the passing on of instructions for assembly via DNA. But with the nanofactory concept, individual nanobots will not reproduce -- indeed, they will carry no instructions for replication; it will literally be impossible.

You can get a good sense of how this all fits together by watching a recently produced short film on productive nanosystems.

Mike Treder

Tags: nanotechnology nanotech nano science technology ethics weblog blog

Comments

[Brett Bellmore]

I think the best answer is that nanosystems could be built so that they *could* evolve, but that there's very little to be gained by doing that, and obvious drawbacks. Most mutations are unfavorable, after all, and it takes a LOT of generations to accuulate much improvement.

I could see a limited form of "evolution" being implemented in nanofactories, though: Machine "learning" of new mechanosynthetic operations, so that the machines would grow more efficient and capable as time went by.

[george]

sir,
any foreign body that enters our body is taken to be an antigen then why doesn the body take the sperm as an antigen and try to act on tat wen its in the females body....

[Brett Bellmore]

Because, topologically, the sperm ISN'T in the woman's body. Among other reasons...

[cdnprodigy]

It will be when I start using better pickup lines...

[Chris Phoenix, CRN]

A point worth repeating is that "nanobots" such as medical devices will not have the ability to reproduce. In biology, each organism has to have come from a similar organism and be able to make more. In engineered nanosystems, each product can and should be an "evolutionary dead end" without any ability or mechanism for reproduction.

Even the kind of "machine learning" that Brett refers to does not require reproduction. A non-mutating nanofactory can make lots of non-reproducing products with various changes, and the products can then be tested and further iterations of designs produced, all using the original design of nanofactory.

There's simply no need for products to make more products. It would be like each car having to tow around an oil refinery, not to mention a mechanic's shop.