The Plan to Feed the World by Hacking Photosynthesis

I was under the impression that nitrogen fixation was still the limiting factor in plant growth. If that's the case, will improving photosynthesis really improve yields?

Would improved photosynthesis be liable to find its way into nearby weeds via horizontal gene transfer? It sounds like a really advantageous modification, and so something that's likely to be picked up on. (Whereas, say, a modification for higher yields wouldn't be, because it wouldn't be advantageous for a weed.)

Are C3 and C4 really simply interchangeable mechanisms? What about CAM, would it be possible to "install a CAM module" into maize to make it drought resistant, though less productive?

Also, but that might not be directly related to your work, the article says "plants already absorb more sunlight than they can use" but still goes on to talk about how they could be more efficient if they used a larger part of the light spectrum. Would it serve any purpose, then?

Monoculture, petrol, chemical agriculture is often produced in a centralized manner, which introduces inefficiencies (with environmental issues) in transport, fertilization, pest control, etc. The current paradigm also requires large amounts of capital to purchase machinery, mined fertilizer, poisonous pesticides, GMOs where seeds can't be saved, etc. This cycle of dependence has caused many small farms to be in financial crisis. Also, don't forget that large scale factory farming is subsidized by the government. Without these subsidizations, large scale factory farming is not viable.

Factory meat production follows the same pattern of monoculture, with disgusting & unhealthy results.

The alternative is to decentralize food production. Small, organic, polyculture farms reduce or eliminate the need to transport food over long distances, import fertilizer, use pesticides. Decentralized food production also enables the farmer to be independent of having to purchase expensive inputs from large multinational corporations.

Small farms are more humane and, similar to startups, freer to innovate.

If we can do more with less, we will use less; less space, less soil augmentation, less everything. That's still a good thing.

Also, the more minimalistic our food plants become, the more alternatives we create for ourselves; if nothing else, space travel and settlement is going to demand as much efficiency in our crops as we can get.

This is partly true, but is not helpful.

The goal is for nobody to go hungry.

By increasing crop yields in the most common staple crops, we have a clear technological engineering problem to solve and various ways to approach it, all achievable. This allows more food to be produced more cheaply in the areas that need it most.

If we refuse to produce more food until the food that is already produced is distributed evenly, many people will die or go hungry because the socio-political problems are very complex. There is no engineering solution to them in sight.

By all means let's try to solve problems of inequality. But let's also try to feed people as soon as possible by whatever means are available.

There are two parts to the answer.

Firstly, some plants have evolved to do photosynthesis more efficiently. One of the projects in the article (the one I work on) is to take a particularly efficient kind of naturally occurring photosynthesis - C4 photosynthesis - and copy it over to species that don't already have it. Sugarcane is an example of a plant with very efficient photosynthesis.

Secondly, there are relatively few evolutionary innovations in photosynthesis for several reasons. The main one is that the global environment is constantly changing - the atmosphere is very different now than it was 1.6 billion years ago when photosynthesis first evolved. The atmosphere started out with no oxygen, so photosynthesis first evolved to handle that scenario. But photosynthesis releases oxygen by breaking up water molecules, so it gradually oxygenated the atmosphere and depleted the carbon dioxide. Modern conditions with very high oxygen and low carbon break a lot of the assumptions that were fixed into the early photosynthetic system, so improvements tend to be workarounds because evolution mainly builds on what is already there.

The thing that humans have that evolution doesn't is the ability to reason about the system. Evolution is like an optimisation algorithm working within constraints, and humans are like software engineers that can change the constraints.

The article gives at least one answer to that question: the over concentration of chlorophyll can be valuable when competing against other plant. This is a case where evolutionary fitness doesn't align with the goals of agriculture.

This is one out of two typical answers which are

1) The fitness function is different, so there would be a trade-of but we don't mind it.

2) The biochemistry needed is very complicated, requiring many steps and is unlikely to be discovered by evolution.

To your last point: Evolution is prone to local maximums. We evolved to think beyond them.

It isn't wise. Every time somebody figure's out how to optimize for one variable, everyone starts optimizing and we end up with great productivity for a brief bit and then the California drought.

We have been lucky so far cause scaling up takes time.

There's also, I fear, a trap here where we believe we understand nutrition, only to discover there are nuances that we still don't have a great handle on.

As an example, there's the emerging posibility that one's gut flora can have a huge impact on overall health - and that certain types of foods can have a negative or positive impact on that gut flora.

Manipulating basic elements could easily lead to our skipping over critical parts of our health needs, with unintended consequences down the line.

Put another way: our dietary habits and nutritional systems developed over tens-to-hundreds of thousands of years. I'm very leery of any theory that we can easily replace that.

I'll use an analogy with "nanorobots":

When I was doing nanotech 10+ years ago, and people asked me when nanorobots were coming, I'd always ask them why they thought nanorobots would be different than, say, bacteria, or viruses? The constraints "at the bottom" (as Feynman would say) are quite high—and we're not really very close to being able to engineer at that scale. Heck, even making a robot the size of a grain of rice is a pretty daunting task, and a nanorobot would be built on a scale that's 10^12 to 10^18 times smaller (the cubic relationship is a real bitch), depending on what you mean by "nanoscale".

Engineering a "thing" to convert basic chemicals and sunlight into food is a pretty significant task. While I would bet that plants are nowhere near close to optimal, they're already solving the problem. Changing & rethinking an existing system is far easier than simply creating one from whole cloth.

If I had to guess, I'd say that "producing food from manipulating basic elements" using a piece of tech entirely designed by human hands, from scratch, is at least a couple of centuries in the future.

If we ever get to tame energy I don't think we'll need to care about water and food. It seems inefficient to tame energy only to then invest more time in food research to give our bodies energy.

I'm pretty confused by your logic, I certainly don't think that it implies, for example, that the reader would be unable to go into education and go down a career path where they would become one of the few experts who are working in this field. It even links to papers and technical articles.

I'm assuming that is because this future WILL be created by experts. That isn't elitist, but if you want to help you will have to put in the time an effort to became an export.