3 ms·

Let’s waste more money on science

23 points by chriskanan 10y ago

Nzen 10y ago
tl;dr we should resist an insistence that federally funded science ought prioritize by free-market principles. Thomas Levenson provides examples of basic research that took long periods, often initially focused on non-remunerative subjects. Ex, Thomas Brock studied boring microbes at Yellowstone, which led to studying unrecognized extremophiles. Those bacteria's enzymes contributed to the use of PCR to identity genes cheaply. (Sure, we can't predict the positive consequences of boring science, but is federally funded money science any less likely to accomplish the same? Distinguished from corporate money science, ala Marlboro's tobacco studies.) <edit for phrasing of first sentence>
- credit_guy 10y ago
  This is a good tl;dr. I have a reservation with this conclusion. For each positive example of basic research that led to some unexpected applications there are hundreds that didn't. One might argue that maybe many of these unapplied pieces of research are still waiting for their time to come. But unfortunately, most scientific results are simply forgotten. One can then continue the argument by pointing out that this is the way of research: one tries a thousand ideas, and only a handful prove fruitful. And in order for this trial and error process to work, you need an environment that promotes interaction, exchange of ideas, and more importantly, risk-taking. And risk-taking is difficult to encourage within the confines of the corporate world, where people live and die by the quarterly earnings. At all other times, I would agree that this is a fair point, but not necessarily today. Today there's a tremendous amount of good quality research that is being done in the corporate world. Deep learning comes to mind. Here's an interesting quote that I found in a book by Stephane Mallat about wavelets: "I cofounded a start-up [...]. It took us some time to learn that in three months good engineering should produce robust algorithms that operate in real time, as opposed to the three years we were used [in academia] to having for writing new ideas with promising perspectives." I hope that people don't share your view that the corporate science is represented by the Marlboro's tobacco studies. I contend that the vast majority of corporate science is good science, and moreover the corporate culture promotes efficiency in research. This comes at a steep cost though, that of severely restricting the movement of ideas and interaction across corporate entities. I don't personally have a solution for this problem, or rather I can see that one solution is the industry consolidation and emergence of corporate giants (Google, Facebook), but I can't say I like this.

23 points by chriskanan 10y ago

Nzen 10y ago

tl;dr we should resist an insistence that federally funded science ought prioritize by free-market principles. Thomas Levenson provides examples of basic research that took long periods, often initially focused on non-remunerative subjects. Ex, Thomas Brock studied boring microbes at Yellowstone, which led to studying unrecognized extremophiles. Those bacteria's enzymes contributed to the use of PCR to identity genes cheaply. (Sure, we can't predict the positive consequences of boring science, but is federally funded money science any less likely to accomplish the same? Distinguished from corporate money science, ala Marlboro's tobacco studies.) <edit for phrasing of first sentence>

credit_guy 10y ago
This is a good tl;dr. I have a reservation with this conclusion. For each positive example of basic research that led to some unexpected applications there are hundreds that didn't. One might argue that maybe many of these unapplied pieces of research are still waiting for their time to come. But unfortunately, most scientific results are simply forgotten. One can then continue the argument by pointing out that this is the way of research: one tries a thousand ideas, and only a handful prove fruitful. And in order for this trial and error process to work, you need an environment that promotes interaction, exchange of ideas, and more importantly, risk-taking. And risk-taking is difficult to encourage within the confines of the corporate world, where people live and die by the quarterly earnings. At all other times, I would agree that this is a fair point, but not necessarily today. Today there's a tremendous amount of good quality research that is being done in the corporate world. Deep learning comes to mind. Here's an interesting quote that I found in a book by Stephane Mallat about wavelets: "I cofounded a start-up [...]. It took us some time to learn that in three months good engineering should produce robust algorithms that operate in real time, as opposed to the three years we were used [in academia] to having for writing new ideas with promising perspectives." I hope that people don't share your view that the corporate science is represented by the Marlboro's tobacco studies. I contend that the vast majority of corporate science is good science, and moreover the corporate culture promotes efficiency in research. This comes at a steep cost though, that of severely restricting the movement of ideas and interaction across corporate entities. I don't personally have a solution for this problem, or rather I can see that one solution is the industry consolidation and emergence of corporate giants (Google, Facebook), but I can't say I like this.