Revision as of 18:30, 14 March 2015

Don't worry. It won't be like that. (image source: DT Eightron ep1)

As mentioned here advanced productive nanosystems like nanofactories are quite unsuitable for food production.

Virtually all molecules that make up our food are very flexible chain molecules often with branches and sometimes with aromatic rings. Those are harder to Mechanosynthesize than simple and stiff diamondoid structures.

One may consider small salt crystals an exception but they get instantly dissolved and do not stay crystalline in the body. Salts are probably best handeled dissolved and sorted with nanopores but they should be handlable mechanosynthetically in solid state too since they do not diffuse at room temperature.

Mechanosynthesis of floppy chain molecule structures

If you carefully read the tooltip paper (referenced on the mechanosynthesis page) you recognize that for the synthesis of stiff diiamondoid carbon structures half rings of several carbon atoms are added. Those half rings behave like floppy chain molecules. Also in the paper longer chains are streched between two tips.

To synthesize the floppy chain molecules one may be able to employ the following trick: First one starts out with at two opposing tooltips that I'll call the tensioning tooltips. With further tooltips one starts to extends the length of the chainmolecule between the two tensioning tooltips. While increasing the length of the chainmolecule the tensioning tooltips must keep the chainmolecule in tension. Extension of the chainmolecule must always happen near the tensioning tooltips where the thermal fluctuations of the chain are smallest. It may be possible to include branches into the molecule into a double ended tree topology. Every branch will need a seperate tensioning tooltip. This will probably require a lot more design effort in comparison to simple mechanosynthesis of diamond and graphitic structures especially if there are many branches which are very close together since the tooltiops start to block there freedome of movement mutually. The classical fat finger challange. Floppy loops of various are another issue.

Synthesis of diamond can be done at room temperature with acceptable error rates. For the synthesis of these more floppy molecules cooling may be necessary. Maybe an ice matrix could be mechanosynthesized. It seems diffecult to drop in mechanosynthesized floppy chain molecules in a controlled way.

In natural systems that is plants and animals nourishment molecules are usually synthesized from larger prebuilt molecule fragments. In an artificial system where smaller groups of atoms are put together more bonds need to be formed that is there is a lot more energy involved. There is more energy turnover. To be efficient mechanosynthetic energy recuperation is necessary. It may be possible to have way lower losses at the same production speed or the same losses at way higher production speed.

Btw: Natural food is already a self replicating technology so APM will not make the cost for food drop in the same radical way as it will for anything factory produced.

The unnaturality of synthesized food

What may be synthesizable ...

Chemical compounds

The basic cemical compounds for food are analogous to the mostly inedible diamondoid compounds (exception e.g. NaCl) in the sense that they form the lowest level basis of products. The main difference here is that one can't get away with just one material for all aplications at all. You may produce pure shugar for mere calories but its common knowledge that you'll get ill from such a diet.

Natural food has millions of chemical compounds in it. There will be a lot of development work needed for each class of compounds. Which is in stark contrast to simple synthesis of diamondoid structures. There you only need to specify the locations for the atoms in high level software and don't have to design new toolpath trajectories. Not to mention even new kinds of tooltips. The sythesized food will consequently for a long time rather be a simple cocktail of nutrients containing only the molecules that one is able to synthesize rather that a food source that is certain to be healthy.

Molecule classes that are of interest are e.g.: shugars, starch, fatty acids, lipids, vitamins, amino acids, nucleotides and many many more

Structure

Recreating the structure of real food on the intermolecular nanoscale is ridiculously diffecult and pretty much unnecessary. Consider the structural complexity of a plant or animal cell (lipid layers cell organelles ...) and the insane amount of chemical compounds involved.

The structure on the micro to makroscale makes the experiencable texture of the food so here patterning the products becomes important. With only a few molecules a lot of textures may be archievable (here the metamaterial approach works - conventional baking as post processing to get something like an chrispy dough?)

It probably won't bee too hard too fool our sense of taste, that is this synthesized food may taste damn good but ....

How good or bad is it for your body

Food that does not contain stuff of which we don't yet know that we need it. Like drinking distilled water. On the short term there will be no consequences but on the long therm there might be deficiency symptoms.

Every single atom of the synthesized food you eat can be potentially be checked on nuclear stability. (Handling of salts ...) (see: isotope sorting)

Degree of necessity

Today a common argument for artificial energy efficient production of food is that we need to be able to continue feeding the growing world population. But there is the overlooked issue that the emergence advanced APM technology is likely to lead to rapid decline of population instead which poses entirely different problems.

Aside that at the point where the products start to look good taste good and are not too unhealthy theres little doubt that it will find widespread use. Nonteless we probably don't want to replace our diet but extend it.