There are 325 million people in the United States. If we give each of them a "basic income" of $1,000 per month (which is bare bones subsistence basic income), the accumulated cost would be $12,000 per year times 325 million people, or $3.9 trillion per year.
The entire US federal budget is already $4.094 trillion. (https://en.wikipedia.org/wiki/2018_United_States_federal_bud...), and once you account for things like defense, interest on the national debt, and Medicare that we'd still have to pay for, it's still looking like we would have to add maybe $3 trillion to the annual budget. (You can get really deep into the weeds on this: for example, even if you replace some or all of Social Security with UBI, some Social Security recipients receive more than $1000 per month, so you would need to top them up to the same levels they're already receiving because of the time-honored principle of The United States Government Does Not Welch On Its Debts).
So here comes the other side of the argument: "you can just raise taxes on rich people to make up for that 3 trillion dollar shortfall!". Yeah, but you can already raise taxes on rich people.
The steelman version of that argument is more like, "even if we give away 4 trillion dollars, we can set up the tax brackets so we get some of it back, so the total cost is less than 4 trillion". OK, cool, let's do some napkin math. It looks like $100,000 is around the 75th percentile of income in the US (https://en.wikipedia.org/wiki/Household_income_in_the_United...), and that's a big round number where people definitely shouldn't be receiving public assistance anymore, so let's make that the break-even point where we tax back all of the UBI we give away in the first place, which means we get to cut our UBI budget by at least 25% (since only 75% of the population gets any), and maybe let's throw in another 12.5% for gradually taxing back some of the UBI for people in the 50th-75th percentiles, too. So our $4 trillion expenditure is now $2.5 trillion, and let's assume we can write off the same trillion dollars as before by canceling out existing welfare and entitlements, so now you're adding a net $1.5 trillion annual expenditure.
That actually seems doable. This would also mean that providing a UBI for ten years would cost $15 trillion. So, OK, let's accept for the sake of argument that the federal government is going to spend $15 trillion over the course of the next ten years to make everyone's lives better. Is this really the best way? Or could you do something like:
* Repair and replace damaged and worn out US infrastructure: $2 trillion (https://www.csmonitor.com/Business/2018/0118/Why-fixing-US-i...), much of which will create jobs and stimulate enough economic activity to pay itself off in the long run.
* A manned mission to Mars: At most, $1 trillion (http://time.com/money/4765718/travel-mars-price-cost-tourism...), much of which will create jobs.
* ITER, the best-funded effort to develop fusion power, costs about $20 billion (http://www.sciencemag.org/news/2016/04/updated-panel-backs-i...). Spend $250 billion on a Apollo/Manhattan Project scale effort to develop low cost energy solutions, either from fusion or other projects. (It'll create jobs!)
* It would cost $10 billion to build a Lofstrom loop, but since the federal government is doing it and we've never done anything like that before, let's 10x that. That's another $100 billion. (And even more jobs!)
And you still have $11.5 trillion left to either spend on even more extravagant government projects or eat a decent chunk out of the $21 trillion national debt (https://www.usgovernmentdebt.us/). And in addition to creating more than enough jobs for everyone who needs one, we get to potentially solve some of the most important long-term problems for human civilization, repair all of our infrastructure, land on goddamn Mars, and knock out half of our national debt.