3 ms·

The amount of fossil fuels it takes to produce each calorie of food is increasing exponentially over time, so many scientists expect that eventually we'll need

The amount of fossil fuels it takes to produce each calorie of food is increasing exponentially over time, so many scientists expect that eventually we'll need to start replacing machines with human labor. While this will probably happen to some extent, I also suspect that AI will soon make urban permaculture viable. E.g. the main reason why all the trees in our cities aren't edible fruit and nut trees is because they make a huge mess and attract rodents when they fall on the sidewalk. But if you were creating a new city today then it would make sense to heavily incorporate edible landscaping, and in twenty years when it's actually at peak productivity we'll have inexpensive autonomous robots to clean up the mess and periodically check the soil for contaminants or whatever.

dsfyu404ed 10y ago
citation very much needed
- Alex3917 10y ago
  https://www.ncbi.nlm.nih.gov/pubmed/12178986 https://www.ncbi.nlm.nih.gov/pubmed/12178986
  - Animats 10y ago
    That's from 1993. Needs an update. Agriculture has been reducing their inputs in recent decades. With precision farming and a lot more data collection, farms use less of everything. Using lots of water and fertilizer is out. It looks like we're not going to run out of energy. Solar and wind have now become Big Solar and Big Wind, and fight it out with Big Oil and Big Gas. (Big Coal went bust.[1]) Electric cars aren't a joke any more. We can still do nuclear if we have to. We're more likely to have metal shortages than energy shortages. [1] https://www.washingtonpost.com/news/energy-environment/wp/2016/04/13/coal-titan-peabody-energy-files-for-bankruptcy/?utm_term=.003968a26f35 https://www.washingtonpost.com/news/energy-environment/wp/20...
    - dredmorbius 10y ago
      The trends over the 20th century generally fit the narrative of exponential growth in energy consumption. Major drivers have been fertiliser (especially nitrogen fixation via Haber-Bosch), farm equipment (especially considering both embodied energy of manufacture plus fuel use), and irrigation costs. Two additional factors are the growth in consumption of both animal products (eggs, dairy, poultry, port, and beef, roughly in that order of increasing feed:food conversion cost), and in the consumption of non-staple foods: fuits and vegetables are vastly less energy dense than corn, wheat, rice, and soybeans, and hence have a higher energy cost:calorie ratio. Vaclav Smil's 1994 Energy in World History has an excellent exposition of these trends generally, though ending 22 years ago. Manfred Weissenbacher's Sources of Power updates the story to 2009. It's also helpful to realise that sub-Saharan Africa still has very little mechanised agriculture (fewer tha 1 million tractors per Weissenbacher), and little fertiliser or pesticide use. Even in the US and Europe, nitrogen fertiliser use didn't take off until the 1950s -- for the first 30 years or so, Haber-Bosch were largely in the munitions business. Since ~1990, energy productivity of ag seems to be increasing (that is, less input energy to output crop) in the US and EU (USDA and DOE have several reports). Global data and graphics are much harder to find, I don't know what the situation is.
  - vonmoltke 10y ago
    That abstract mentions nothing about exponentially-growing energy requirements. It's also an abstract for a paper from a forum sponsored by Negative Population Growth, a non-profit devoted to encouraging a reduction of the world's population for sustainability reasons. I am talking this with the same salt I take a climate change paper put out by energy company scientists.
philipkglass 10y ago
The muscles of humans and other mammals are only about 20% efficient at turning food into mechanical work. Overall efficiency is even worse when you consider resting metabolism during non-work times, such as when people are asleep. That profligate waste is happening on top of the inefficiency of using plants to produce human-edible "fuel" for workers; less than 5% of incoming solar energy gets stored as something humans can eat, or a bit better if you include parts that ruminants can eat. Motors running from battery-backed solar power can do much better than that. Cheap solar modules made of multicrystalline silicon can convert 16% of incoming solar energy to electricity, and after going through a battery and a motor you can still get 80% of the original electricity back as mechanical work. A solar farm "feeding" electrical machinery is 1-2 orders of magnitude more productive on an areal basis than a plot of land feeding manual laborers with human foods. Additionally, the solar farm doesn't need weeding, irrigation, fertilizer, or fertile soil. A reversion to muscle power would be expected following a massive societal collapse, but it would not be a rational plan for a high-tech society to increase its efficiency and sustainability.

dsfyu404ed 10y ago

citation very much needed

Alex3917 10y ago
https://www.ncbi.nlm.nih.gov/pubmed/12178986 https://www.ncbi.nlm.nih.gov/pubmed/12178986
- Animats 10y ago
  That's from 1993. Needs an update. Agriculture has been reducing their inputs in recent decades. With precision farming and a lot more data collection, farms use less of everything. Using lots of water and fertilizer is out. It looks like we're not going to run out of energy. Solar and wind have now become Big Solar and Big Wind, and fight it out with Big Oil and Big Gas. (Big Coal went bust.[1]) Electric cars aren't a joke any more. We can still do nuclear if we have to. We're more likely to have metal shortages than energy shortages. [1] https://www.washingtonpost.com/news/energy-environment/wp/2016/04/13/coal-titan-peabody-energy-files-for-bankruptcy/?utm_term=.003968a26f35 https://www.washingtonpost.com/news/energy-environment/wp/20...
  - dredmorbius 10y ago
    The trends over the 20th century generally fit the narrative of exponential growth in energy consumption. Major drivers have been fertiliser (especially nitrogen fixation via Haber-Bosch), farm equipment (especially considering both embodied energy of manufacture plus fuel use), and irrigation costs. Two additional factors are the growth in consumption of both animal products (eggs, dairy, poultry, port, and beef, roughly in that order of increasing feed:food conversion cost), and in the consumption of non-staple foods: fuits and vegetables are vastly less energy dense than corn, wheat, rice, and soybeans, and hence have a higher energy cost:calorie ratio. Vaclav Smil's 1994 Energy in World History has an excellent exposition of these trends generally, though ending 22 years ago. Manfred Weissenbacher's Sources of Power updates the story to 2009. It's also helpful to realise that sub-Saharan Africa still has very little mechanised agriculture (fewer tha 1 million tractors per Weissenbacher), and little fertiliser or pesticide use. Even in the US and Europe, nitrogen fertiliser use didn't take off until the 1950s -- for the first 30 years or so, Haber-Bosch were largely in the munitions business. Since ~1990, energy productivity of ag seems to be increasing (that is, less input energy to output crop) in the US and EU (USDA and DOE have several reports). Global data and graphics are much harder to find, I don't know what the situation is.
- vonmoltke 10y ago
  That abstract mentions nothing about exponentially-growing energy requirements. It's also an abstract for a paper from a forum sponsored by Negative Population Growth, a non-profit devoted to encouraging a reduction of the world's population for sustainability reasons. I am talking this with the same salt I take a climate change paper put out by energy company scientists.

philipkglass 10y ago

The muscles of humans and other mammals are only about 20% efficient at turning food into mechanical work. Overall efficiency is even worse when you consider resting metabolism during non-work times, such as when people are asleep. That profligate waste is happening on top of the inefficiency of using plants to produce human-edible "fuel" for workers; less than 5% of incoming solar energy gets stored as something humans can eat, or a bit better if you include parts that ruminants can eat. Motors running from battery-backed solar power can do much better than that. Cheap solar modules made of multicrystalline silicon can convert 16% of incoming solar energy to electricity, and after going through a battery and a motor you can still get 80% of the original electricity back as mechanical work. A solar farm "feeding" electrical machinery is 1-2 orders of magnitude more productive on an areal basis than a plot of land feeding manual laborers with human foods. Additionally, the solar farm doesn't need weeding, irrigation, fertilizer, or fertile soil. A reversion to muscle power would be expected following a massive societal collapse, but it would not be a rational plan for a high-tech society to increase its efficiency and sustainability.