Are we headed for a world in which most of the meat we consume was grown in a vat rather than in an animal? This is a fairly high-stakes question (pun intended). We have a growing population, we are already using most of the arable land available, and we are pushing the efficiency of agriculture. There is still some technological head room, mostly with GMO technology, to improve yields further. But we are already use more land than is optimal to feed ourselves, as well as a lot of water, and consuming a lot of energy, which has a carbon footprint. Anywhere we can achieve efficiency can have a huge impact, therefore.

The animal product industry is the focus of a lot of attention by environmentalists because of its inherent inefficiency. There are different ways to look at this. Cattle consume about 25 calories for each one they produce, pork is about 15, chicken 6, and fish is close to 1:1. But we can also look at calories produced per acre of land, in which case the comparison is not as clear.

One way to look at the issue is calories produced per acre of land. Here beef has the worst ratio, with 1.1 million calories produced per acre (I assume this is per year, although it is not explicitly stated in the linked reference). Potatoes are very efficient at 17.8 million calories per acre. But soybeans are also pretty inefficient at 2.1, while pork is more efficient than soybeans at 3.5. The ratio of efficiency between potatoes and soybeans is about the same as the ratio between wheat (6.4) and beef.

Some types of meat are more land-efficient than some plants, but yes, overall, plants are more efficient. The picture is also complicated because animals can produce high-quality nutrition and can use land not suitable for growing plant-based food. Simplistic comparison therefore breakdown when you look at a more complete picture. But I think it is scientifically non-controversial to say that the typical Western diet includes too much meat, and if we cut down it would improve the overall efficiency of our food production in terms of land, water, and carbon. This is a low-tech solution that is likely optimal for health and the environment.

Another low-tech solution is to incorporate more insects into our food chain. This can be in the form of highly processed insect protein and carbohydrates, like cricket powder. This way you are not consuming whole insects. You can bake bread or make pasta out of cricket powder (the products I see available are only partly cricket powder and partly wheat flour), and you cannot detect the insect origin, although there might be a slight difference in flavor. You can also feed insects to animals, making meat production more efficient and environmentally friendly.

What about high-tech solutions. This is where cultured meat comes in, something I have been interested in for years. In 2016 I was pretty optimistic about lab grown meat, and now I am still optimistic but more tempered. I think I fell for a common futurist mistake of overestimating short term progress, but in my defense the industry was also overly optimistic. The problem is in estimating how much and how fast the technology will advance. Lab grown meat involves using cell cultures from the muscle of different animals, like cows, and growing them in bioreactors – essentially large vats. There is a clear advantage to this technique in terms of reduced land, water, and energy use. There is also the potential to tweak the health profile of the meat itself.

But the big remaining question is cost. Bioreactors are expensive to build and operate. The optimistic projections of cultured meat are dependent upon a dramatic reduction in cost through improvements in technology. They are also dependent upon the ability to scale up production. Both of these are problematic assumptions, and so far have proven overly optimistic. A more skeptical analysis finds that there may be inherent limits on the technology of cultured meat that will make it expensive for the foreseeable future.

For example, even a very large and expensive bioreactor facility could produce in a year what a similarly priced meat processing plant can produce in a week. About 5,000 such facilities would be necessary to produce the current output of the meat industry. That’s not impossible, just very expensive.

Still, there are companies claiming significant reduction in the cost per pound of cultured meat. An Israeli company, Future Meat, claims it can produce cultured chicken at $7.70 per pound, about twice as expensive as meat from chickens. Given that the price started at about $30,000 per pound, it does not seem unreasonable to think that another 50% reduction is possible.  A cultured meat burger would now set you back about $9.80. That’s an expensive burger, but within the range you might pay at a restaurant rather than fast food.

So while projections 6 years ago were a bit optimistic, critics have been to pessimistic. There is a tendency to overestimate short term progress (things are usually harder than they first appear and unanticipated hurdles crop up) but also to underestimate long term progress (technological feedback loops can produce exponential progress). For example, by now we were supposed to have fully self-driving cars, but that last 5% of error reduction was a lot harder than imagined. But we’ll get there, perhaps just delayed by about a decade. It’s no surprise that lab-grown meat is taking longer than proponents originally projected, but it seems like we are getting there. It may take another decade, even two, before the price is down far enough and production volume is high enough for cultured meat to have a significant impact, but it is likely coming.

In the meantime, eat more plants, and get used to the idea of eating insects.