As I sometimes do, I listened carefully to a fairly recent podcast on this topic and took some notes. I include the link to this podcast and my notes below. You can scan my notes and decide it if is worth your time to listen to the podcast.
If you do, you already have a rough idea of what is in the podcast, and what is discussed. You have a sort of "guide" to the podcast.
If you do not want to bother listening, you can maybe get something out of the notes, at least.
This podcast, and some similar podcasts, might be folded into my recent draft essays on R&D which I have posted here on Thinkspot.
Science is in trouble and it worries me.
https://www.youtube.com/watch?v=QtxjatbVb7M
Physicist Sabine Hossenfelder explains in this podcast.
Innovation is slowing, research productivity is declining, scientific work is becoming more disruptive. In this video I summarize what we know about the problem and what possible causes have been proposed. I also explain why this matters so much to me.
00:00 Intro
00:33 Numbers
06:33 Causes
10:32 Speculations
16:25 Bullsh1t Research
22:06 Epilogue
-Scientific progress is slowing down and most of what is published in academia is bullsh1t.
-number of scientists have increased with time, in absolute numbers (now more than 8.8 million) and as a share of the population
-number of scientific publications are growing
-physics publication volume doubles every 18 years
-electrical engineering literature doubles every 9 years
-all the effort we put into science has fewer and fewer results
-the most obvious measure is productivity, comparing inputs with outputs, economically
-total factor productivity TFP is one metric
-the input is labor, capital and technologies
-the output is goods and services that you can sell
-in 2016 a group of economists looked at this in a paper called 'Are Ideas Getting Harder To Find?' by Bloom, Jones, Reenen and Webb from the National Bureau of Economic Research in Cambridge Massachusetts (working paper 23782)
http://www.nber.org/papers/w23782
-There has been a steady growth in TFP in the US of about 5% a year in spite of a huge growth in the number of researchers
-also new drugs approved is dropping and crop yields are stable in spite of huge and growing number of new researchers
-paper 'Combinations of technology in US patents 1926-2009: a weakening base for US innovation' by Matthew S. Clancy, Economics of Innovation and New Technology, 2018, Vol 27, No. 8. 770-785 found that US patents made less novel connections among technological constituents since the 1950s
-by this measure, patent novelty has been going down since the 1960s in the US
-thee same was reported in another paper 'Invention as a combinatorial process: evidence from US patents' by Youn, Strumsky, Bettencourt and Lobo, 2015, J. R. Soc. Interface 12 20150272 using US patent records from 1790 to 2010, and found narrowness of inventions in the US has increased
-'Are 'flow of ideas' and 'research productivity' in secular decline?' by Cauwels and Sornette, published in 2022 in Technological Forecasting & Social Change, found that research productivity (measured by top researchers per community) has steeply decreased since the 1960s
-last year a study appeared in Nature analyzing 45 million papers published worldwide; 'Papers and patents are becoming less disruptive over time' by Park, Leahy and Funk, published in 2023. They used the measure of how many papers were made redundant by a given publication and found that the number of disruptive ideas has gone down in many areas of science with time.
-in 2005 Jonathon Huebner published 'A possible declining trend for worldwide innovation' in Technological Forecasting & Social Change, 72 (2005), 980-986 found that the rate of innovation peaked in 1873 and is now rapidly declining (defined as number of technological events divided by the number of people)
-these studies all find that in research we are making more efforts for less in return
-this seems contradictory since things seem to be changing so fast now
-But AI, quantum computers, brain implants, nuclear fusion are based on breakthroughs that are now decades old
-What are the possible causes?
-R&D funding in the US has not decreased since the moon landing and has remained at roughly the same level (although the federal share of R&D funding has dropped), but this is not just an American phenomenon
-US government share of US R&D support is roughly 19% now, and has been approximately stable at this level for decades (although it was 67% in 1960). Corporate share of US R&D support has now grown to 74% (up from about 31% in 1960).
-hiring more people has partly made up for the lower productivity per researcher
-if companies did not get something out of R&D, they wouldn't do R&D
-maybe the low hanging fruit has already been "picked" in R&D? Would this explain this phenomenon?
-John Horgan in his book 'The End of Science' said there are no big breakthroughs left
-Sabine thinks this is unlikely to be true, and the slowdown is probably mostly because of the way we have organized R&D in recent years
-in the US and the EU most research is privately funded
-most basic research, with a long time horizon (that one expects to deliver the breakthroughs), is publicly funded and is in academia
-during WWII and immediately after there was tremendous pressure and a shared sense of purpose
-research and long term research was different and better in the US and the EU until about the 1970s or so
-there is a growing problem with fraud in STEM R&D, but it is not yet a major problem
-we know science is slowing down and it is impeding societal progress
-Is the origin of the problem in academia?
-venture capitalists notice this too and many wonder if it is too much bureaucracy
-Elon Musk has a lot of stories about how bureaucracy stands in the way of progress at his R&D-based companies
-Peter Thiel thinks the large projects like Apollo created large lethargic bureaucracies that were or are politicized
-The 'publish or perish' precept values quantity over quality
-Academic research is becoming risk averse because there is so much competition for funding and jobs; this explains the lack of novelty and disruptiveness in research results because no one wants to take a chance. They want to only work on a "sure thing", that is, low risk mainstream research, because they want to get published and get a job and so on.
-European Research Council wants to fund only transformative research or high risk, high payoff research, but then they also fund the same old stuff instead
-it has become a mark of pride among some scientists in academia to proclaim that science is just boring
-many produce bullsh*t research, like the book 'Bullsh*t Jobs; The Rise of Pointless Work and What We Can Do About It' by David Graeber; the definition of 'bullsh*t jobs' (and bullsh*t research) is, 'stuff we would be better off without'
-this erodes the trust in science, but most of it is invisible and the average person never hears or sees anything about it, so it is limited to small circle of people so far, yet
-the public will soon start to ask questions
-Patrick Collison of Stripe and his collaborators offered "fast grants" to researchers without a lot of bureaucracy during the pandemic
-Collison did a survey among those receiving grants from him, and asked them how satisified they were with their research; 78% of those surveyed and funded said that if they could spend their research dollars on anything they wanted, they would change the theme of their research a lot; that is, they would work on something else, if only they could. This was a survey of people doing biomedical research, but Sabine also surveyed physicist about 20 years ago, and got similar results. Most people in R&D would change the topic of their research efforts if they could.
-Sabine interprets this to suggest that most people in academia do not do the research they want to do, they do not do what they consider the most worthwhile research, only 'BS' research, and they know it
-of course, every so often, something that looks worthless turns out to be good for something; in German one has the expression that every so often, a blind chicken sometimes finds a kernel of grain
-How does one find worthwhile research, research that is "worth the money"? Sabine does not know, and this has her worried.
-we do not currently have the technological progress we need to protect ourselves from a supervolcano or an asteroid impact or a large solar flare. We need to protect our species. Technological progress depends on scientific progress, and technological progress is stalling because we are ignoring what is going wrong in academia.
-Sabine does not want her children to die because the technical publishing company Elsevier keeps publishing junk.
Filter By:
