> So once you take BI received - taxes paid, people that earn more outside income get less from the BI program than those that don't. And we are back in the same situation where a welfare program creates disincentives for work.

If you fund a UBI program by, say, eliminating preferential rates for capital income, and adding supplemental brackets above the current maximum brackets, and tie the level of the UBI to (because the former actually simplifies reporting in a way which prevents doing this exactly rather than by estimates) an estimate of the revenue resulting from those changes, there's no actual disincentive to work compared to the status quo ante due to the UBI except for very high income workers(this isn't no additional disincentive compared to those in other welfare programs from the UBI, its no disincentive at all.)

There's a disincentive compared to the status quo ante to generating income from long-term capital gains, but the practical effect of that is probably only in the upper end of the distribution (in the lower and middle ranges, the relative comparative disincentive to earning that form of income is probably more than offset by the having more surplus income to invest.)