DeclineInInnovation

ThoughtStorms Wiki

Context: ProblemsWithScience, OnInnovation

Various people express concern about a decline in innovation and productivity.

They are concerned with conformity and group-think in academia. "Woke ideology" Etc.

What I think :

Firstly I'm somewhat sceptical about this decline and how it's measured. It seems to me that

a) compared to much of human history we are still making quite a respectable amount of innovation.
b) measuring innovation quantitatively is going to be pretty hard (and based on very subjective criteria) anyway.
c) in particular I don't think patents are a good proxy for innovation. For the obvious reason that they aren't evidence of new ideas, just of legal manipulation. Furthermore some areas of innovation lend themselves to patents more than others. And some communities embrace patents more than others. (Thankfully AFAICT, and to their credit, the AI people aren't out there trying to patent every new machine learning architecture. They are publishing it openly. This doesn't mean no innovation is going on. Quite the opposite. But again patents aren't a great proxy for measuring it.)
d) we have undoubtedly picked a LOT of the lower-hanging fruit in the last three centuries. There are physical limits in things like material science. Eg. once you've gone to lithium, the lightest metal, for batteries, you aren't going to find any lighter ones. Similarly, new particle accelerators, new telescopes, new chip-making techniques, or language models all need to be ever larger (or smaller) or consume more energy to make progress over the current state of the art. This does not mean that there aren't still great and revolutionary ideas to be had cheap. But they may not be as plentiful as they were 100 years ago.
e) much cutting edge research is extremely abstract and complicated. It takes more time and effort than ever for any individual researcher to get to a state within which they can contribute to pushing the boundaries. So while we are throwing more people than ever at research, those people need to be ever more exceptional to be effective.

Do I think AI and AutomatingScience will help? Yes, I do. I think a new generation of AIs building models, crunching data and inventing new hypotheses will supercharge our existing scientific practice. It will certainly fill in the blanks and comprehensively scan most of the places we need scanned.

Will this actually create a sense of "faster innovation"? I'm not so sure. I think the perception of the rate of innovation is already very subjective. It depends how familiar you are with various existing trends. How much you speculate about what is available to discover. Even what you like as a discovery. How much a new thing feels like an innovative discovery rather than an unfortunate junk trend.

In the sphere of music, I often argue that AutoTune is somewhat analogous to the invention / discovery of distortion for electric guitars. Many people love the effects on music that distortion brought to it. For over three decades, from The Kinks to Pink Floyd to My Bloody Valentine, the artistic possibilities of distortion of electric guitars have proven a spur to musical innovation. Many people who would violently agree with me on that, will nevertheless see Autotune as mere tawdry abomination. One that has no artistic merit and produces no interesting innovation in music.

Many people's reactions to the new waves of GenerativeAI are similar. For me, the output of ChatGPT is a marvel of our current technological research. Someone else sees merely the hallucinations it makes, the fact it has imbibed quantities of other people's writing without permission, and the horrific damage it might yet do to our culture, as evidence that it is nothing but a dangerous scam. (AIIsFakeAndSucks)

Here's something else I think.

Neither commercial nor cultural innovation are great proxies for scientific or technological innovation. While often new science or technologies can lead to commercial returns, this is not always the case. And sometimes a commercial or cultural innovation does not depend on any particular scientific or technological innovation. Even though I am tempted towards MusicalTechnologicalDeterminism much of the time, I recognise that sometimes cultural change is just cultural change.

I do nevertheless have a theory for the decline in growth in productivity in the West over the last 40-50 years. I believe it's to do with the move from manufacturing industry to "service industries".

My theory is that "productivity" is easier to improve in manufacturing where you are trying to maximise the number of widgets rolling off the production line. And harder in "services" which are often selling hours of human time. You can't actually sell more human hours per human hour because it's the same unit.

Therefore in all the economies that have deindustrialized and moved to service economies, productivity growth is going to slow down. Regardless of other factors.

And that means you can't take productivity growth as a good proxy for innovation or the ROI on science.

Contrast :

Backlinks (3 items)