3 ms·
I have a feeling that the creators might be wrongly assuming that we could evolve plants to thrive in extreme monocultures (no other plant or animal life). I su
by karol 5y ago
I have a feeling that the creators might be wrongly assuming that we could evolve plants to thrive in extreme monocultures (no other plant or animal life). I suspect these efforts will be met with diminishing returns of a) seeing lower yields because plants don't need to compete for resources b) really hard to go beyond 99% of desired plant because of energy consumption c) weeds adapting to fool the AI and resemble desired plants in phenotype and d) people who tread on the soil to weed by hand interact with the soil in a different way than a robot on wheels.
- f6v 5y agoHow is this different from people pulling the weeds? They could have evolved to fool us already. I don’t see a huge difference since image recognition is something computers can do on par with humans.
- igammarays 5y ago> since image recognition is something computers can do on par with humans No way. Computer image recognition is nowhere close to human recognition in real-world contexts. Still full of errors and bugs.
- imtringued 5y agoThere are worse things that you can do to the soil, things like tilling it.
- contingencies 5y agoIt is very reasonable revisit the assumption that fighting nature in this way is a reasonable course of action. We can not solve our problems with the same level of thinking that created them. - Einstein There are some promising reports of enhanced agricultural outcomes from reduced inputs based on alternative strategies such as seedballs, dense intercropping, crop inter-rotation, less intense land use and higher biodiversity. Major issues with such approaches seem to be homogeneity (required by large scale distributors) and difficulties with autonomous harvesting or increased labour (increased yields are no good if you can't harvest them efficiently).