The internet, on which he has a lot to say, has had enormous benefits, but a striking amount of online activity is free and internet businesses create few new jobs (and displace lots of others). The result is growth in utility without much of a contribution to GDP, which would be fine except that countries and people have bills to pay, on things like health care, pensions, and government debt. Complicating matters is the fact that the fastest growing contributors to measured GDP—the government, education, and health sectors—deliver returns that are very difficult to measure. This suggests, he says, that rich world GDPs are likely overstated; we're poorer than we thought.
Well this is terrible... GDP isn't rising because too many things are free! So a profession emerges and builds a measure that some might argue was wrong-headed to start with. As things change, it becomes clear that it's getting less and less useful measure. And what conclusion does this lead the commentators to? (No, not that we should just forget that GDP was ever invented.) They conclude things must be getting worse because lots of things are happening that don't register on their measuring instruments.
How terrible is it that so many valuable things are increasingly freely available? Well if you've got a vested interest in charting the gap between rich and poor, it's pretty bad, evidently.