> If you model a back-of-the-envelope sustainable civilisation, you'll find we're nowhere near any sort of fundamental population limit.

How did you model it? I'm curious about the assumptions you made and how you came up with your conclusion.

I lost the envelope (it was a while ago (and also not a literal envelope)), but I'm pretty sure I just drew a circle somewhere reasonably habitable in Europe, filled it with a decent population density to roughly non-capital city population, placed enough high-density farmland around it to produce double the necessary food (in crops), put in eight radial train tracks and some roads leading to those train tracks, surrounded the whole thing by enough 20%-efficiency solar panels to power the city (peak 300W per person, at 10% electricity losses) then multiplied up to nine billion people.

I assumed all in-city transport was manual or electric (and hence covered by the solar panels), all trains and food transport was diesel (even though electric trains are realistic), that the roofs of the city were covered in 20% efficiency solar panels (at 0% loss, due to proximity), that all long-distance transport was done with cargo ships (no aeroplanes) and there was a tenth the shipping of what we've actually got.

The vast majority of the planet was completely untouched by my hypothetical civilisation. I know I missed out a lot of stuff (e.g. how goods get from ports to people and back), but I don't think my conclusion was that far out – especially since “tile the world with solar panels” isn't an efficient renewables strategy.