7 ms·

People say you need to let scientists do useless exploration in order to promote innovation. I think that is true to a degree, but definitely should not be the

by mjfl 7y ago

People say you need to let scientists do useless exploration in order to promote innovation. I think that is true to a degree, but definitely should not be the spirit of any policy where innovation is the goal. Many of the scientific innovations that pushed the 20th century forward were in fact purposefully done. The Von Neumann architecture was not invented to have fun, it was invented to aim artillery and design nukes. Similarly, much of the tech infrastructure in Silicon Valley descends directly from radio engineers coming out of WWII theater. Bardeen, Brittain, and Shockley didn't invent the solid state transistor because it would be interesting, they invented it because it would allow miniaturization of vacuum tube computers. The laser was invented for telecommunications purposes. There is now an entire field of "quantum electronics" describing the theory behind them. Even going back further, the invention of the steam engine preceded the development of thermodynamics, which initially sought to describe the limitations of these engines. IF "science leads technology", then one would expect steam engines to be rationally designed from the results of thermodynamics. The opposite is true. I think there are enough examples of scientific fields emerging from technological innovation that "uselessness" should not be considered correlated with innovation.

bsder 7y ago
> I think there are enough examples of scientific fields emerging from technological innovation that "uselessness" should not be considered correlated with innovation. I was going to cite lots of "useless" stuff that suddenly became really important when some "blocking technology" got removed, but "useless" is the wrong problem and phrasing. The problem really is: "How do you compensate someone who tackles a big problem but winds up making no progress or even being wrong?"
- mjfl 7y ago
  From personal experience, pressure and flexibility. The investigator should be put under pressure to produce results quickly justifying further time investment, in addition there needs to be flexibility in the form of alternative projects they can tackle that have a high probability of working. But I think it's crucial that they be allowed to place themselves under that pressure, if they truly believe they can accomplish whatever big thing it is. And if they have a really good reason that can be written up, that itself is a result that justifies further time investment. I think current science eliminates the possibility of that pressure, and focuses on the high probability projects. I think places like Google X actually don't put enough pressure on, though I don't really know that organization very well.
BurningFrog 7y ago
I think you're mostly talking about engineering, not science. Engineering always has a practical goal, and for that field you're right.
nine_k 7y ago
The von Neumann architecture became possible because Church and Turing wrote their highly theoretical, mathematical works. Lasers became possible because people like Fresnel studied the properties of a highly impractical phenomenon, coherent light, while other people discovered another short-lived curiosity, inverse energy levels population. Many such works had to be done decades before any engineering applications, or a prospect thereof. Science is when you study the literally unknown, including no known practical applications. When you study small pockets of unknown in a generally understood and practically fertile field, it's engineering.
- mjfl 7y ago
  > The von Neumann architecture became possible because Church and Turing wrote their highly theoretical, mathematical works. I don't know about others but Von Neumann architecture is much more described by Church and Turings work than derived from it, like steam engines are described by thermodynamics. Neither invention was actually dependent on the theory. Science often derives from engineering, not the other way around. Get over it.
  - marmaduke 7y ago
    Science and engineering are a cycle on a spectrum, you don't get one without the other.
  - Karrot_Kream 7y ago
    > I don't know about others but Von Neumann architecture is much more described by Church and Turings work than derived from it, like steam engines are described by thermodynamics. Neither invention was actually dependent on the theory. Huh? Turing's paper on the halting problem and Universal Turing Machines was released in 1936. Von Neumann, Eckert, and Mauchly's memo was not released until Jan 1944. There is no way that Von Neumann's architecture is described by Turing's work, given the 8 year difference between them! > I don't know about others I'm puzzled by how you can make such a strong claim about the nature of innovation without precisely drilling down into the evidence backing your theory.
voxl 7y ago
You're literally just wrong: https://www.youtube.com/watch?v=y2I4E_UINRo https://www.youtube.com/watch?v=y2I4E_UINRo
rwallace 7y ago
Right. However, the inventions you mention, were not expected to be useful the same year. The problem today is not that there is a demand to focus work on potentially useful areas; that's fine as far as it goes. The problem is that there is too much demand that work bear fruit in the short term, whereas in fact we have picked most of the fruit from the branches currently in reach, to mix metaphors, and we need to go back and do some more long-term investment.

by mjfl 7y ago

bsder 7y ago

> I think there are enough examples of scientific fields emerging from technological innovation that "uselessness" should not be considered correlated with innovation. I was going to cite lots of "useless" stuff that suddenly became really important when some "blocking technology" got removed, but "useless" is the wrong problem and phrasing. The problem really is: "How do you compensate someone who tackles a big problem but winds up making no progress or even being wrong?"

mjfl 7y ago
From personal experience, pressure and flexibility. The investigator should be put under pressure to produce results quickly justifying further time investment, in addition there needs to be flexibility in the form of alternative projects they can tackle that have a high probability of working. But I think it's crucial that they be allowed to place themselves under that pressure, if they truly believe they can accomplish whatever big thing it is. And if they have a really good reason that can be written up, that itself is a result that justifies further time investment. I think current science eliminates the possibility of that pressure, and focuses on the high probability projects. I think places like Google X actually don't put enough pressure on, though I don't really know that organization very well.

BurningFrog 7y ago

I think you're mostly talking about engineering, not science. Engineering always has a practical goal, and for that field you're right.

nine_k 7y ago

The von Neumann architecture became possible because Church and Turing wrote their highly theoretical, mathematical works. Lasers became possible because people like Fresnel studied the properties of a highly impractical phenomenon, coherent light, while other people discovered another short-lived curiosity, inverse energy levels population. Many such works had to be done decades before any engineering applications, or a prospect thereof. Science is when you study the literally unknown, including no known practical applications. When you study small pockets of unknown in a generally understood and practically fertile field, it's engineering.

mjfl 7y ago
> The von Neumann architecture became possible because Church and Turing wrote their highly theoretical, mathematical works. I don't know about others but Von Neumann architecture is much more described by Church and Turings work than derived from it, like steam engines are described by thermodynamics. Neither invention was actually dependent on the theory. Science often derives from engineering, not the other way around. Get over it.
- marmaduke 7y ago
  Science and engineering are a cycle on a spectrum, you don't get one without the other.
- Karrot_Kream 7y ago
  > I don't know about others but Von Neumann architecture is much more described by Church and Turings work than derived from it, like steam engines are described by thermodynamics. Neither invention was actually dependent on the theory. Huh? Turing's paper on the halting problem and Universal Turing Machines was released in 1936. Von Neumann, Eckert, and Mauchly's memo was not released until Jan 1944. There is no way that Von Neumann's architecture is described by Turing's work, given the 8 year difference between them! > I don't know about others I'm puzzled by how you can make such a strong claim about the nature of innovation without precisely drilling down into the evidence backing your theory.

voxl 7y ago

You're literally just wrong: https://www.youtube.com/watch?v=y2I4E_UINRo https://www.youtube.com/watch?v=y2I4E_UINRo

rwallace 7y ago

Right. However, the inventions you mention, were not expected to be useful the same year. The problem today is not that there is a demand to focus work on potentially useful areas; that's fine as far as it goes. The problem is that there is too much demand that work bear fruit in the short term, whereas in fact we have picked most of the fruit from the branches currently in reach, to mix metaphors, and we need to go back and do some more long-term investment.