Checking I understand this right: the paper’s contention is that more R&D, whilst still producing productivity gains, is invalidating gains made from other R@D? e.g. everyone developing their own LLM. It’s an interesting idea, but why would that be happening?
If the OECD is to be believed, the elephant in the room is that China negated the value of western R&D by about $500B per year since 2010.
That’s why being a fast follower is so valuable, you get everyone else to waste money on the wrong ways to build things. It also causes R&D to show much worse returns.
I do not, and have never, understood why absolute R&D investment is compared to relative growth rate in this literature. It would make sense to compare R&D spending as a percentage of GDP to relative growth, or to compare absolute R&D spending to absolute per-year growth, but I remain mystified by the comparison of absolute R&D spending with relative growth.
They show R&D is effective and R&D spending is up, and conclude obsolescence must be the reason this is not reflected in productivity pretty much by process of elimination. However there is an alternative - that for reasons unrelated to R&D, productivity is actually being driven down, and ever increasing R&D output is necessary just to maintain current levels of productivity.
In particular, they are looking at US manufacturing. While this is a diverse industry, it's clear that in many subfields there is quite a bit of saturation. When everyone has a car, and cars last longer and longer, the need for new cars goes down. Once you get to a point where your industrial capacity can provide enough to satisfy demand, further R&D only reduces the costs of satisfying that demand, not increased output, and in some cases improvements to product quality may even further reduce demand. In the US, light vehicle sales peaked in 2000, and while the numbers dipped during various market downturns, they keep coming back to roughly the same asymptote. The numbers are even more striking if you break it down further - the annual demand for personal vehicles has fallen by a factor of 4 since 1965, and by a factor of 3 since 2000, the difference being made up by increased commercial vehicle demand. Looking at other industries like steel paints a similar picture.
They are specifically looking at R&D in Manufacturing. I think there you can make the case (as they do) that one new innovation can erase the productivity benefits of a prior innovation.
I am writing a book about this topic. I started my career in 1978 at Bell Labs and worked in 3 different startups after that. After 45 years in R&D, I have recently retired. So many times, the inspiration for new inventions we worked on came from unexpected sources; the arts, culture, music, history and many other sources. And I said we on purpose because rarely did a new invention come from one person, it was almost always from collaboration on a team. My conclusion is that invention so often comes from a team of well rounded people with knowledge in many areas and the ability to work in a team. I wonder if the decline in the productivity in R&D comes from a decline in these attributes?
This would seem to be a direct corollary to the red queen hypothesis applied in the context corporations instead of species. That is, in a competitive environment you have to keep spending R&D dollars to stay in the same relative market position because everyone else around is spending as well. However the paper talks about productivity of the individual firm and aggregate productivity (presumably across the whole economy). Therefore I think that the red queen may not be whole story, because firms should still be getting more efficient (more productive) even if they can't capture that value due to competition, the production possibilities frontier should be growing because we need less capital to accomplish the same tasks, leaving more for other things. However it seems that this is not the case? So what the paper seems to mean by "increased rates of obsolescence" is that there is so much churn within organizations that they can't actually get something implemented in a way that actually allows them to capitalize on the potential increased productivity? That sounds like a complexity wall, but I feel like I'm feel like I'm missing something.
R & D is about the only thing you have if you want to make something out of nothing.
Most other forms of value-added activity need to start with something of value to begin with.
The closer that that "nothing" can be brought to zero, the greater the leverage by comparison until it wipes the floor with everything else.
It wasn't so bad until all the MBA's came along, they have nothing like the equations that are needed to figure this out when the data is not numbers yet. Regular non-degreed business operators used to be so much more advanced in mathematical intuition regardless.
The researchers stayed as talented, plus got better technology, the founding giants of leadership lasted as long as they could but were replaced by midgets, and here we are.
You can recycle any old idea into a brand-new-and-patentable idea by adding "with AI" to the end. For example, "assessing value for used cars" is an old idea, but "assessing value for used cars WITH AI" is a new idea
"These findings suggest that R&D has become more effective at finding productivity-enhancing ideas, but these ideas may also render rivals’ technologies obsolete, making innovations more transient. Because of obsolescence, rising R&D does not necessarily mean rising aggregate productivity growth."
I Don't Unserstand that. Innovative should increase productivity. If i deploy a new technology i should see productivity gains. Does that Not mean we see should se productivity gains regardless of how fast a technology becomes obsolete? Maybe technologies do not get adopted that fast or the productivity gains do not justify the Investment?
14 comments
[ 6.4 ms ] story [ 38.8 ms ] threadThat’s why being a fast follower is so valuable, you get everyone else to waste money on the wrong ways to build things. It also causes R&D to show much worse returns.
In particular, they are looking at US manufacturing. While this is a diverse industry, it's clear that in many subfields there is quite a bit of saturation. When everyone has a car, and cars last longer and longer, the need for new cars goes down. Once you get to a point where your industrial capacity can provide enough to satisfy demand, further R&D only reduces the costs of satisfying that demand, not increased output, and in some cases improvements to product quality may even further reduce demand. In the US, light vehicle sales peaked in 2000, and while the numbers dipped during various market downturns, they keep coming back to roughly the same asymptote. The numbers are even more striking if you break it down further - the annual demand for personal vehicles has fallen by a factor of 4 since 1965, and by a factor of 3 since 2000, the difference being made up by increased commercial vehicle demand. Looking at other industries like steel paints a similar picture.
Most other forms of value-added activity need to start with something of value to begin with.
The closer that that "nothing" can be brought to zero, the greater the leverage by comparison until it wipes the floor with everything else.
It wasn't so bad until all the MBA's came along, they have nothing like the equations that are needed to figure this out when the data is not numbers yet. Regular non-degreed business operators used to be so much more advanced in mathematical intuition regardless.
The researchers stayed as talented, plus got better technology, the founding giants of leadership lasted as long as they could but were replaced by midgets, and here we are.
Evading patent minefields, however ...
Especially now that AI can continuously scan for infringement, ideas are getting harder to use.
I Don't Unserstand that. Innovative should increase productivity. If i deploy a new technology i should see productivity gains. Does that Not mean we see should se productivity gains regardless of how fast a technology becomes obsolete? Maybe technologies do not get adopted that fast or the productivity gains do not justify the Investment?