Modern-day inventors—even those in the league of Steve Jobs—will have a tough time measuring up to the productivity of the Thomas Edisons of the past.
That's because big ideas are getting harder and harder to find, and innovations have become increasingly massive and costly endeavors, according to new research from economists at the Stanford Institute for Economic Policy Research. As a result, tremendous continual increases in research and development will be needed to sustain even today's low rate of economic growth.Nicholas Bloom, a SIEPR senior fellow and co-author of the forthcoming paper, contends that so many game-changing inventions have appeared since World War II that it's become increasingly difficult to come up with the next big idea.
[...] Turning its focus to publicly traded companies, the study found a fraction of firms where research productivity—as measured by growth in sales, market capitalization, employment and revenue-per-worker productivity—grew decade-over-decade since 1980. But overall, more than 85 percent of the firms showed steady, rapid declines in productivity while their spending in R&D rose. The analysis found research productivity for firms fell, on average, about 10 percent per year, and it would take 15 times more researchers today than it did 30 years ago to produce the same rate of economic growth.
https://phys.org/news/2017-06-big-ideas-harder.html
[Source]: https://siepr.stanford.edu/news/productivity-ideas-hard-to-find
[Paper]: Are Ideas Getting Harder to Find?
Do you think that innovative ideas are hard to find ??
(Score: 0) by Anonymous Coward on Tuesday June 06 2017, @06:22PM (1 child)
I know there are some who hate when I bring this up... but once again look at figure 2. The blue curve (their index of research efficiency) is basically the inverse of NHST adoption. This is the mass confusion predicted by Ronald Fisher:
Fisher, R N (1958). "The Nature of Probability". Centennial Review. 2: 261–274
(Score: 1) by khallow on Wednesday June 07 2017, @01:29PM
I believe you misinterpret the loyal opposition here. The point of NHST is to do science in situations where you have a pile of data and don't know enough to do the usual hypothesis and model building. No one is arguing that p-hacking and other failure modes of NHST don't happen, but the technique has a legitimate use.
What's relevant here is that NHST is supposed to be a temporary technique. You mine your data, find possible correlations, and build models from there. You shouldn't use NHST forever, because both of its flaws - the p-hacking trap and its natural inefficiency, but also because you supposedly have models to test now. The growing use of NHST over the past century indicates that there are a number of fields that simply aren't progressing on to model building, instead stalling at the NHST stage.
I will agree that if you're merely interested in the appearance of doing science rather than actually making progress, then NHST is a great technique for looking busy. So heavy, long term use of NHST is a warning sign that we are doing things seriously wrong.