You can't say that because we're assuming ceteris paribus and we've already defined our control as the productivity of food on a given plot of land.
Our food production per plot of land has gone up; or, our required size of land to produce the same amount of food has gone down.
We can't say for sure if one person's profits will go up or down, at least not without additional information about the particularities and price elasticity stats of the market, because both their costs and their revenue have changed. But markets change all the time, people (and farmers especially) know they have to change produce from time to time, depending on what's profitable. Overall, though, lower costs are a good thing. Always. That's exactly what's happening here.
Societies where most people are in agriculture tend to be societies where most people are poor, and this is a causal effect: Their costs are so high for farming they can't afford to have industry elsewhere. Reducing costs means fewer people have to be in agriculture, and this is good.
This isn't even a feedback loop, although sometimes people will make a similar argument around other phenomenon assuming all feedback loops must be a positive feedback loop that never decays, incorrectly reaching the conclusion the economy will eventually collapse. (E.g. "Prices went up, therefore the cost of producing/refining oil/gas will go up, therefore the cost of producing many products will go up, until all products cost infinity!")