9 ms·

How Should the U.S. Fund Research and Development?

18 points by tokenadult 11y ago

hartator 11y ago
I think the question is bit biased. I think we should ask first if the idea of the fund itself is a good idea.
- rhizome 11y ago
  You might refer to the long-standing grant systems at DOE, NEA, SBA, and so on.
- brownbat 11y ago
  > I think the question is bit biased. I think we should ask first if the idea of the fund itself is a good idea. That's literally the first issue the article discusses in detail. It's in bright red lettering in bold font.
- drjesusphd 11y ago
  Since the times of Newton, science has always been funded by governments. Funding agencies are as much part of the real-world scientific process as The Hypothesis.
  - WalterBright 11y ago
    Not exclusively. Einstein wasn't.
    - drjesusphd 11y ago
      He was a government employee in 1905 when he published special relativity, Brownian motion, and the photoelectric effect. This work was done while he was a student studying under a government-funded scientist. He was a full professor by the time he released general relativity.
      - WalterBright 11y ago
        He was a patent examiner. Not a government funded researcher.
brownbat 11y ago
A lot of scientific studies (some funded by grants) are never tested through replication, even though that's a core part of the process. Replication is the unsexy grunt work that makes science reliable, but it usually doesn't make anyone famous (or necessarily even get them tenure), because it's not bleeding edge work creating or advancing new theories. Maybe the government should treat this like a market failure and generously fund replication studies. Planet Money on the replication issue: http://www.npr.org/sections/money/2016/01/15/463237871/episode-677-the-experiment-experiment http://www.npr.org/sections/money/2016/01/15/463237871/episo... Replication crisis (wikipedia): https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis
- vph 11y ago
  To be fair, scientists by and large understand the need of replication, and evaluation in general. It's just in certain fields (e.g. psychology) or circumstances with human subjects, it's very expensive or even infeasible to have well controlled repeated experiments.
  - Godel_unicode 11y ago
    If you don't have well controlled, repeatable experiments (a.k.a. the scientific method) are you actually doing science? I think we should come up with another word for these types of one-off studies.
    - gumby 11y ago
      Many fields advance without solid well controlled, repeatable experiments, at least through various stages of development: cosmology, geology, most of medicine, philosophy, science of mind, etc. You don't consider those all "science"? There's more to science than Bacon.
      - freyr 11y ago
        Those fields have scientific and non-scientific components. The non-scientific parts are not science.
vph 11y ago
I believe that whoever is in charge of federal funding in R&D will have to be extremely smart and balanced in his views because R&D plays a great role in the economy and future of this country. The right approach is a balanced between fundamental and applied research. And it's not just balanced but which areas to invest money into.
- Godel_unicode 11y ago
  I think that the government would be better served to spend money researching fundamental science, that seems to me to be the great shortfall of industry. Intel, Westinghouse, biotech in general, and Dow do a pretty good job on the applied piece IMHO.
  - vph 11y ago
    >Intel, Westinghouse, biotech in general, and Dow do a pretty good job on the applied piece IMHO. Private companies are in a great position to do applied research, but they are often slow or unwilling to distribute the research results for the benefit of everyone.
    - Godel_unicode 11y ago
      They actually do a good job of that as well, it's just that they do it in the form of patents which means that you have to pay to benefit from their work.
joeblow9999 11y ago
by letting free individuals decide what they want to fund?
WalterBright 11y ago
> computer networking I've been in computing long enough to know that the first thing anyone does with 2 computers is to connect them together. The notion that nobody would have thought of it without Arpanet is absurd. People invented networking protocols ALL THE TIME before the internet. FidoNET, BIX, Prodigy, MCImail, AOL, RBBS, CompuServe, Kermit, The $25 Network, etc. Even I designed a scheme (I never got around to implementing it), even calling it "web" (!).

18 points by tokenadult 11y ago

hartator 11y ago

I think the question is bit biased. I think we should ask first if the idea of the fund itself is a good idea.

rhizome 11y ago
You might refer to the long-standing grant systems at DOE, NEA, SBA, and so on.
brownbat 11y ago
> I think the question is bit biased. I think we should ask first if the idea of the fund itself is a good idea. That's literally the first issue the article discusses in detail. It's in bright red lettering in bold font.
drjesusphd 11y ago
Since the times of Newton, science has always been funded by governments. Funding agencies are as much part of the real-world scientific process as The Hypothesis.
- WalterBright 11y ago
  Not exclusively. Einstein wasn't.
  - drjesusphd 11y ago
    He was a government employee in 1905 when he published special relativity, Brownian motion, and the photoelectric effect. This work was done while he was a student studying under a government-funded scientist. He was a full professor by the time he released general relativity.
    - WalterBright 11y ago
      He was a patent examiner. Not a government funded researcher.

brownbat 11y ago

A lot of scientific studies (some funded by grants) are never tested through replication, even though that's a core part of the process. Replication is the unsexy grunt work that makes science reliable, but it usually doesn't make anyone famous (or necessarily even get them tenure), because it's not bleeding edge work creating or advancing new theories. Maybe the government should treat this like a market failure and generously fund replication studies. Planet Money on the replication issue: http://www.npr.org/sections/money/2016/01/15/463237871/episode-677-the-experiment-experiment http://www.npr.org/sections/money/2016/01/15/463237871/episo... Replication crisis (wikipedia): https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis

vph 11y ago
To be fair, scientists by and large understand the need of replication, and evaluation in general. It's just in certain fields (e.g. psychology) or circumstances with human subjects, it's very expensive or even infeasible to have well controlled repeated experiments.
- Godel_unicode 11y ago
  If you don't have well controlled, repeatable experiments (a.k.a. the scientific method) are you actually doing science? I think we should come up with another word for these types of one-off studies.
  - gumby 11y ago
    Many fields advance without solid well controlled, repeatable experiments, at least through various stages of development: cosmology, geology, most of medicine, philosophy, science of mind, etc. You don't consider those all "science"? There's more to science than Bacon.
    - freyr 11y ago
      Those fields have scientific and non-scientific components. The non-scientific parts are not science.

vph 11y ago

I believe that whoever is in charge of federal funding in R&D will have to be extremely smart and balanced in his views because R&D plays a great role in the economy and future of this country. The right approach is a balanced between fundamental and applied research. And it's not just balanced but which areas to invest money into.

Godel_unicode 11y ago
I think that the government would be better served to spend money researching fundamental science, that seems to me to be the great shortfall of industry. Intel, Westinghouse, biotech in general, and Dow do a pretty good job on the applied piece IMHO.
- vph 11y ago
  >Intel, Westinghouse, biotech in general, and Dow do a pretty good job on the applied piece IMHO. Private companies are in a great position to do applied research, but they are often slow or unwilling to distribute the research results for the benefit of everyone.
  - Godel_unicode 11y ago
    They actually do a good job of that as well, it's just that they do it in the form of patents which means that you have to pay to benefit from their work.

joeblow9999 11y ago

by letting free individuals decide what they want to fund?

WalterBright 11y ago

> computer networking I've been in computing long enough to know that the first thing anyone does with 2 computers is to connect them together. The notion that nobody would have thought of it without Arpanet is absurd. People invented networking protocols ALL THE TIME before the internet. FidoNET, BIX, Prodigy, MCImail, AOL, RBBS, CompuServe, Kermit, The $25 Network, etc. Even I designed a scheme (I never got around to implementing it), even calling it "web" (!).