In that case We don't have enough work to do. Automating agriculture and distributing produce for free ( or at low margin ) is the last nail in to the coffin. This all is just dopamine, real work needs bombastic goals.
We destroy more than enough food to end world hunger, but actually distributing that food to the people that need it would undermine the authority of lots of well-armed organizations.
De-weaponizing food distribution would do a heck of a lot more than further reducing the percentage of the population that works in agriculture.
Put another way, 10% of the US works in agriculture-related fields (including restaurants, food processing and textile manufacturing), but only 1.2% works on farms:
https://www.ers.usda.gov/data-products/chart-gallery/gallery...
Driving farm jobs from 1.2% to 0% isn’t going to move food costs or availability very much, and it’d barely be perceptible to the economy at large.
If you could remove the somebody from the chain then the entire cost to produce food at that point would be "fiction", ie it would be possible to make it free. But as long as human labor is involved in the production it cannot be free.
I’d argue that production labor is close enough to free for all practical purposes.
Focusing on distribution or environmental impact would have significantly more practical impact.