Fighting over forecasts is meaningless; fighting over 10 year projections - like those from the CBO or CRFB - is ridiculous. Here's why. Financial models do one thing: they extrapolate linearly based on simple assumptions. By definition, they can't anticipate or predict discontinuities. The two most substantial events in the past 25 years were 1) the global financial crisis, and 2) covid, both unprecedented discontinuities. The first added an extra 7 trillion in debt over a decade; the second, an extra 10 trillion over 5 years. Neither event could have been predicted by the CBO and no budget ever could account for these. This means that 17 trillion - or almost half of total US debt of 37 trillion - is the result of just two outlier events which were never planned or expected by Congress or accountants. For better or worse, our present situation and our debt crisis is shaped by just these two events as it is by all the other political issues that have been beaten to death. There are other discontinuities such as recessions. Note the CBO never forecasts a recession, even if one is virtually guaranteed every 5-7 years, and a recession - more than any single expense category - leads to fiscal stimulus and is a periodic debt bomb. That's why CBO forecasts are and always have been a joke at best, and something only politicians or the media pretend to take seriously. So please don't. One other point: just like long-term projections fail to account for negative discontinuities they also can't account for positive ones. AI could be one such positive event, and if it spurs the kind of productivity revolution it has been hyped for, the resulting economic and revenue growth would be unprecedented. Nobody expected it, but Clinton managed to balance the budget and even got a surplus between 1997-2001 in large part thanks to the internet revolution which sparked a historic wealth boom. Imagine what a true AI revolution could do.