Excellent framing on how intercropping systems like the milpa solve core problems modern ag actively creates.The economics around robot armies versus monoculture machinery is particularly sharp, especially the part about diversifying risk across dozens of crops instead of betting everything on one. What I found compelling is the cascade effect once soil fertility starts compounding instead of depleting,the whole model flips. I've seen permaculture setups where biodiversity alone crushes pest pressure without any chemical interventions. Scaling this with autonomous harvestng could legit crack food scarcity.
What an amazing reply. Thank you so much for taking the time to read the essay with such an incisive eye for detail. You hit the nail on the head, multiple times. Cascade effects is exactly where all of this is headed, and I think the economics of autonomous harvesting make this nearly inevitable.
Hmm. Fun to read, I guess, in the way it would have been fun to read Tom Swift a generation ago, but pretty silly. Funny the Aztec way of agriculture didn’t catch on. Probably Monsanto sacrificed it.
By the time Monsanto was founded in 1901 mechanized monocrop agriculture was already well underway as a practice, just look at the Bonanza Farms of the 1870s - 1890s. Monsanto provided the chemical and eventual genetic software layer to the hardware layer of industrial agriculture. So Monsanto had no role in sacrificing milpa technology — it had already been steamrolled by the industrial age.
Ok, so here is one quibble. Industrial age/mechanized farming has depleted land of nutrients, minerals etc. fair enough. but no cost to replace those lost resources? Sure legumes can provide nitrogen, but where do we replace phosphorus potash zinc sulfur boron etc.? Maybe that happens in the transition period, but every harvest will deplete nutrients. Cascade effects will sol e the problem? And oh, the Monsanto comment was a joke .
By the way, giving a timeline doesn’t answer the question of why we don’t practice Aztec ag today.
I'll reply by moving through your comments in reverse.
I attempted —perhaps unsuccessfully— to address your question of why we don't use milpa tech today in the essay:
"The issue is with our current state of ‘automated’ agriculture. An enormous machine designed to harvest 32 identical rows of corn (The Calmer 3215) can’t do anything with a milpa, except maybe destroy it. A milpa is nuanced and agile; monoculture, blunt and straightforward. Thus, monoculture requires blunt and straightforward machines..."
It's an issue of scale: we can't scale the milpa tech with industrial age machines. We **need** autonomous humanoid robots in order to scale milpa tech.
You're right that harvests remove nutrients—thermodynamics doesn't care about our utopian dreams. But food forests address this through:
(1) deep-rooted perennials mining subsoil minerals and trace elements like zinc (fruit/nut trees like apple, walnut, pecan; deep-rooted herbs like comfrey, chicory),
(2) nitrogen-fixing species (leguminous trees like black locust, honey locust; shrubs like Siberian pea shrub, autumn olive),
(3) mycorrhizal networks that access locked-up phosphorus,
The key difference: monoculture depletes nutrients 13-69x faster than regeneration while destroying the biological systems that regenerate them.
Food forests, on the other hand, actively rebuild those biological systems, allowing regeneration to keep pace with—or even exceed—depletion. We know this works because traditional polyculture systems like the milpa sustained populations for centuries without synthetic inputs, and decades of data on regenerative agriculture show soil carbon and mineral content increasing over time.
So yes, with good design, we can rebuild everything that's been depleted. Will food forests need occasional mineral supplementation? Probably. But we're talking rock dust every few years, not synthetic fertilizer every season. And with robotic monitoring, we can detect and address deficiencies before they become problems. The transition period will certainly have hurdles, but once established, a well-designed food forest should be net neutral or positive on most nutrients.
Excellent framing on how intercropping systems like the milpa solve core problems modern ag actively creates.The economics around robot armies versus monoculture machinery is particularly sharp, especially the part about diversifying risk across dozens of crops instead of betting everything on one. What I found compelling is the cascade effect once soil fertility starts compounding instead of depleting,the whole model flips. I've seen permaculture setups where biodiversity alone crushes pest pressure without any chemical interventions. Scaling this with autonomous harvestng could legit crack food scarcity.
What an amazing reply. Thank you so much for taking the time to read the essay with such an incisive eye for detail. You hit the nail on the head, multiple times. Cascade effects is exactly where all of this is headed, and I think the economics of autonomous harvesting make this nearly inevitable.
Hmm. Fun to read, I guess, in the way it would have been fun to read Tom Swift a generation ago, but pretty silly. Funny the Aztec way of agriculture didn’t catch on. Probably Monsanto sacrificed it.
What was silly in your view?
And why is it funny the Milpa didn't catch on?
By the time Monsanto was founded in 1901 mechanized monocrop agriculture was already well underway as a practice, just look at the Bonanza Farms of the 1870s - 1890s. Monsanto provided the chemical and eventual genetic software layer to the hardware layer of industrial agriculture. So Monsanto had no role in sacrificing milpa technology — it had already been steamrolled by the industrial age.
Ok, so here is one quibble. Industrial age/mechanized farming has depleted land of nutrients, minerals etc. fair enough. but no cost to replace those lost resources? Sure legumes can provide nitrogen, but where do we replace phosphorus potash zinc sulfur boron etc.? Maybe that happens in the transition period, but every harvest will deplete nutrients. Cascade effects will sol e the problem? And oh, the Monsanto comment was a joke .
By the way, giving a timeline doesn’t answer the question of why we don’t practice Aztec ag today.
I'll reply by moving through your comments in reverse.
I attempted —perhaps unsuccessfully— to address your question of why we don't use milpa tech today in the essay:
"The issue is with our current state of ‘automated’ agriculture. An enormous machine designed to harvest 32 identical rows of corn (The Calmer 3215) can’t do anything with a milpa, except maybe destroy it. A milpa is nuanced and agile; monoculture, blunt and straightforward. Thus, monoculture requires blunt and straightforward machines..."
It's an issue of scale: we can't scale the milpa tech with industrial age machines. We **need** autonomous humanoid robots in order to scale milpa tech.
You're right that harvests remove nutrients—thermodynamics doesn't care about our utopian dreams. But food forests address this through:
(1) deep-rooted perennials mining subsoil minerals and trace elements like zinc (fruit/nut trees like apple, walnut, pecan; deep-rooted herbs like comfrey, chicory),
(2) nitrogen-fixing species (leguminous trees like black locust, honey locust; shrubs like Siberian pea shrub, autumn olive),
(3) mycorrhizal networks that access locked-up phosphorus,
(4) dynamic accumulator plants (comfrey accumulates potassium, phosphorus, calcium; dandelion accumulates potassium, copper, boron; yarrow accumulates phosphorus, copper, potassium; stinging nettle accumulates iron, sulfur, potassium),
(5) continuous organic matter cycling, and
(6) biological weathering of parent rock.
The key difference: monoculture depletes nutrients 13-69x faster than regeneration while destroying the biological systems that regenerate them.
Food forests, on the other hand, actively rebuild those biological systems, allowing regeneration to keep pace with—or even exceed—depletion. We know this works because traditional polyculture systems like the milpa sustained populations for centuries without synthetic inputs, and decades of data on regenerative agriculture show soil carbon and mineral content increasing over time.
So yes, with good design, we can rebuild everything that's been depleted. Will food forests need occasional mineral supplementation? Probably. But we're talking rock dust every few years, not synthetic fertilizer every season. And with robotic monitoring, we can detect and address deficiencies before they become problems. The transition period will certainly have hurdles, but once established, a well-designed food forest should be net neutral or positive on most nutrients.
Love this essay!
I’ve been in the nutrition and human health optimization field for several decades now, and I’m thrilled with the potential you’re articulating here!
Let’s go!
Thank you for taking the time to read it! I really appreciate it.