Former permaculturist here - who at one point was working on this very idea, as it happens. I think permaculture appeals to a lot of programmers, because it's

Former permaculturist here - who at one point was working on this very idea, as it happens. I think permaculture appeals to a lot of programmers, because it's almost like programming with plants. It presents itself as "these plants relate to each other and support each other and you can compose them into guilds that will provide you food with minimal maintenance!" It's complex system design. The problem is that it's not actually based in hard science. We don't know if the nitrogen produced by nitrogen fixers is available to plants around them. It might only become available once those plants die back. The studies that support the idea of dynamic accumulators measured nutrient amounts in the leaves - not the nutrient amounts in the soil as the plants grew and died back. Most of the "companions" are based on folk tales, not research. The whole permaculture design system is the classic "anecdote is not data" fallacy turned into a design system. It's based on individual observation, not careful data collection and experimentation. The criticisms of the reductionist approach that science often falls into are well taken, as are the criticisms of industry's failure to account for unintended consequences when applying science. But that doesn't mean the alternative is to go back to a pre-scientific approach of individual naturalists observing the world and making proclamations based on their observations. Which - having spent five years embedded in the permaculture community and philosophy - is exactly what permaculture is. This failures of anecdotal observation means that permaculture in practice often boils down to the very thing it was meant to change: humans thinking they can engineer and control nature to fit their vision, instead of learning from nature and fitting their vision to the natural realities. Permaculture doesn't actually end up designing systems that work with nature most of the time, it designs novel ecosystems with non-native plants that could have massive unintended consequences. There's a lot to be said for the core concept of farming in nature's image, and I think Mark Shepherd is potentially on to something. We do need to explore whether we could redesign our farm systems to be habitat and feed us at the same time. Food forestry as an idea is a really promising one. But the suburbs need to go away if we're to live sustainably, not be turned into a bunch of experimental gardens. And any agroforestry system needs to be built to fit into the native ecosystem it is a part of, using native plants. And we need to base our approach to designing a permanent culture and regenerative agriculture in science and careful research, not individual observation.