The problems are not fundamentally of local scope. Consider for example the coal mining towns that have been left in a state of ruin by the shift away from coal. There is no local prosperity to fund local aid there.
The US doesn't have a shortage of land or housing. Rather, the concentration of jobs in big cities induces localized shortages. UBI would actually make it feasible to move to areas with lower living costs, decentralizing demand for housing and pushing rent down in big cities while stimulating the economies of less densely-populated regions.
At the end of the presentation the presented mentions that this is structurally identical to loopy belief propagation... Isn't that a big issue, since they inherit many of its tractability issues with regards to training and inference? Modern DL models are far too interconnected for inference to be tractable in general, so the best we can hope for is that we can make simplifying assumptions that make loopy belief propagation feasible.
As a side note, when modern compilers optimize abstract syntax trees, I'm pretty sure they do operations that are similar to the message-passing algorithm described. And they work great, albeit for specialized purposes.
The problem is much bigger than the existence of these gig economy services. The deeper issue is that demand for unskilled labor has become weak enough that, for many people, this is their best option relative to their skill set. Working conditions are terrible, but I don't think it would be a good thing for the world if Uber and the rest of the gig economy services magically disappeared one day: it would simply put the drivers into a financial tailspin with no real chance of escape. Moreover, Uber and Lyft are in a race to the bottom in terms of prices, and neither can unilaterally raise driver wages without pushing riders to the other app. The global economic rat race to drive down prices is unrelenting.
I suspect that universal basic income will become extremely popular in the public view, bordering on necessity, to offset the rise in inequality in the United States.
This was built through incremental edits from a Python proof-of-concept, which you can see here:
https://github.com/yeahpython/ascii-renderer/blob/master/fun...
It doesn't have any real physics, but reading through it might help reveal how the rendering works.
I'd highly recommend learning some improvisational theatre. It requires a lot of the same skills that being charismatic does. Taking an improv class in college was a revelatory experience to me because it showed me exactly how I had been sabotaging my own social interactions. If you can't take a class or a workshop, try to read "Improv Wisdom".
"Two people enter into a series of transactions, and both have the same probability of winning some amount of wealth from the other, just as in a free-market transaction."
I don't really think this is a great description of free market transactions at all. Do you regret the decision anywhere close to 50% of the time when you buy groceries, pay rent, or buy new boots? Does the grocer, landlord or shoe store owner regret it anywhere close to 50% if the time? The answer is no in both cases.
I think the reasoning is that optimizing the training performance is "easy", whereas optimizing the test performance is "hard". If you can guarantee that test performance will be close to training performance, then optimizing the test performance becomes "easy".