If instead we invest in US productivity destroying activities such as dismantling factories, price gouging, speculation, mass diss-information and distraction machines, net loss corporations, and indefinite warfare. Then you would expect productivity to fall eventually.
Today we produce 1,000,000x more videos/memes/gifs/content 50 years ago and we also consume more of them.
But, those numbers don't show up at all in our productivity measures.
Why would watching "Gone with the Wind" 30 years ago more productive because we paid $10 than watching the same "Gone with the Wind" on a streaming service today (amortized at probably $0.05)
The ability to make sufficient for mass market consumption quality combinations of visual, audio and textual content has gone from being purview of real professionals who do this as their day jobs to something literally anyone with a smartphone, an option and 5min of free time can do.
That is a change to human communication as revolutionary as the telegraph or the telephone. There is definitely some value there. Lord knows who'll capture that value.
Truth is nobody knows how to value those things in terms of money, to be honest I'm not even sure they can be valued the same way as we used to value physical things that got built during a physical production process.
I've seen this problem partially addressed by slightly leftish-leaning economists such as Mariana Mazzucato in "The Value of Everything" [1] and I'm sure there may be others like her, but afaik mainstream economics is still ignoring the issue, I think mainstream economists don't even acknowledge this as being an issue.
[1] https://marianamazzucato.com/books/the-value-of-everything
Personally I think we can use Vickrey Auctions for this problem. Vickrey Auctions are able to give not only price discovery, but also value discovery. Everything digital has infinite supplies, that we artificially constrain so as to pump prices. If instead of using price discovery, we use value discovery, the price can stay at 0, so long as a certain (limited) amount of people are faced with a choice that would have them go without.
So for example perhaps we sample 50,000 people and say 75% of people keep access, say when they vote. If you bid in the top 75%, you keep access for the period, and you pay the value that the bottom person that kept access bid. That money is then not used to pay for the IP, some other tax will fund the IP or Open Source Software or Meme, that 25% will then go to all the voters for providing their values. So the average bidder will pay nothing, the below average bidder will lose something they value low, but gain, and the highest bidders will fund that. We then take some other collection and sample another 50,000 people, until we find out the value of all of these digital items to people.
If you do the subsidization process correctly, I think the idea of IP owners holding their supply back would go away, as releasing it to all would always provide more value than holding it back. I think at the very least this would be an obvious answer to Open Source Software, and this would easily fund things like Linux and Firefox. I think this would stretch further to research, and colleges would now be funded through this mechanism, and the professors who currently waste enormous amounts of time doing grant writing would be able to be productive. Thus the whole issue of journals goes away. Even things like ads, people would bid on their service less because of them, so there would be no fundamental reason to include them as it would reduce the subsidy.
There is this weirdly persistent leftist idea of demanding absolute precognition with values for things which are determined in the future and utterly irrelevant and unknown. If I buy coffee beans from locals of Cape Horn growers for my use at the local prices I am helping by supporting local farmers. If I roast them and sell them overseas for 100x the cost I am somehow harming them now.
The Marxist derived notions of value and exploitation are downright certifiably insane.
Of course the problem is that any theory of value like the labor theory has massive holes as well. That's why the current subjective value theory is not really a theory: it's not wrong, it's a placeholder.
30 years ago people valued watching "Gone with the Wind" at $10 and today at $0.05. To serve that, the distributor did a proportional amount of work 30 years ago and today (because it's easier to distribute today). Separately, it would take some effort to produce something equally engaging today with so many entertainment options to choose from.