Instead of subsidizing agriculture at the production side, you can combine farm subsidies, food stamps, and ubi into one consumer credit, designed to persuade people to consume mostly free food over luxury food.
Whats the next thing it would make sense to subsidize for everyone without creating some kind of cyclical purchasing power loop?
As economists say "inflation is a monetary phenomenon". So if you pay for UBI by printing money, then of course it would increase inflation. But Friedman's original proposal and most others pay for it by having the rich pay for the poor through taxes and the middle class pay for themselves.
And on the demand side, the poor in America are generally obtaining food and housing somehow. The food's not nutritious and the housing might be a shelter, but only a small fraction of the homeless let alone the working poor are actually sleeping outside.
On the housing side, proponents hope that UBI will encourage people to move out of the expensive areas where jobs are to areas that are cheaper to live in. So a good argument could be made that housing costs will go down.
On the food side, food is a commodity. The price of a commodity is the marginal cost. So as long as the producer costs are similar between producers, demand has very little effect on the price of food.
It would increase inflation in this example not because the government is printing money, but because the source of the money (the "rich") are not in the market for low-end housing, but the recipients of the taxes are.
However, for rent, there is real scarcity in supply AND limited ability to raise production. I expect that UBI will be an inflation ratchet for rent, unless it is paired with a policy that somehow increases housing supply.
UBI could help more inland cities become more attractive to business as their economies would grow with UBI, but have more land to build housing on.
I don't want to be in the location that passes UBT only to see it fail wildly, but we need that data. I would certainly argue an economy the size of the U.S. or China (as if...) shouldn't experiment without better data.
Personally, I seriously doubt UBT will fail wildly, and I want to see it move forward, responsibly.