4 ms·

Throwing money at the problem may not necessarily speed up research or give us more results.

by kiba 15y ago

pyre 15y ago
How else would you pay for top-calibur talent to sold the problem?
- bilbo0s 15y ago
  I believe what kiba is intimating is that no matter how much you pay... some drugs and diseases are just hard to come up with solutions for. To give an example everyone is fairly familiar with... even if you paid USD100 Million per research position to get a cancer vaccine, I'm not certain you would get it any faster than if you paid USD200000 per position. Or USD100000. It's just REALLY hard to come up with.
  - melling 15y ago
    But what I was suggesting was putting more people on the problem, not offering more per position. In fact, the research done in India would cost less than in the US.
- kiba 15y ago
  Figure out a way to speed up the research process itself. We can't afford to wait 10 to 20 years. Many lives will be lost during those time period. Yes, research take time. Yes, medical research is extraordinary hard. It still doesn't excuse our slowness. We need to look at fighting diseases as a war, not a scientific exploration project. Science is good, but solutions are better. We don't need to understand in extreme details how cancer work, we just need a way to defeat cancer.
  - melling 15y ago
    Add more cores or make the existing ones faster? I said add more cores. If you can think of a way to make each core faster, that's great. In the meantime, getting more people on the problem should help. [Update] Please include include links that discuss the "better" method that you are describing. It sounds like you just want to "reinvent/disrupt" the process without actually having a plan.
    - kiba 15y ago
      You should instead dedicate newer core into process optimization. We want to find a cure in 9 years, then 8 years, 7 years, 6 years, and so on. With each passing year, medical research should be progressing even faster. We don't want to wait ten years for a slightly higher chance of a cure appearing. We want a higher chance of a cure in a much shorter time period. Please include include links that discuss the "better" method that you are describing. It sounds like you just want to "reinvent/disrupt" the process without actually having a plan. It's not a "plan", or a "method". It's a goal. You could start with a small. Figure out what people are doing that should be done by robots or computers. What mistakes people are making that delay the whole research process? Are the bureaucrats holding up the approval for no good reason? How much time you spent fighting politics versus getting things done? The key insight is this: Speeds absolutely matter in medicine. One more day you wait for a cure, somebody die. If speed is not one of your nation's research goal, it should be.
      - SoftwareMaven 15y ago
        This sounds like a very naive view of biological systems. The complexity of the most complex thing computer science has created (I would argue that is the Internet) is still trivially simple compared to the brain. What that means is the search space is huge, much larger than any space I've dealt with in computer science. However, there may be something to what you are saying (I don't know enough about medical research to know whether they are already doing this or not). In public key cryptography, you have a huge key space to try if you want to brute force a key (2^64 for a 64 bit key, unsurprisingly). However, you can break a 64 bit key in about 2^40 operations if you use the right technique [1]. Are we spending enough time looking for those techniques? I certainly hope so. There are a lot of smart people in medicine. 1. http://en.wikipedia.org/wiki/Meet-in-the-middle_attack http://en.wikipedia.org/wiki/Meet-in-the-middle_attack
        kiba 15y ago
        This sounds like a very naive view of biological systems. The complexity of the most complex thing computer science has created (I would argue that is the Internet) is still trivially simple compared to the brain. I am not naive about the difficulty of the problem. Are we spending enough time looking for those techniques? We aren't necessarily looking for better techniques all the time. It could simply be streamlining a step or two in order to make medical research loop tighter. It could mean eliminating a process or two that shouldn't be running. Whatever the case, our researchers should be thinking about novel solutions to the problem instead of mechanically running the lab. The world fastest machine might not be able to make an intractable problem tractable, but it shouldn't be a bottleneck.
- pradocchia 15y ago
  That's a fashionable yet very narrow framing of the matter. If you insist on this formulation, then indeed many many problems are insoluble.

Throwing money at the problem may not necessarily speed up research or give us more results.

by kiba 15y ago

Throwing money at the problem may not necessarily speed up research or give us more results.

pyre 15y ago

How else would you pay for top-calibur talent to sold the problem?

bilbo0s 15y ago
I believe what kiba is intimating is that no matter how much you pay... some drugs and diseases are just hard to come up with solutions for. To give an example everyone is fairly familiar with... even if you paid USD100 Million per research position to get a cancer vaccine, I'm not certain you would get it any faster than if you paid USD200000 per position. Or USD100000. It's just REALLY hard to come up with.
- melling 15y ago
  But what I was suggesting was putting more people on the problem, not offering more per position. In fact, the research done in India would cost less than in the US.
kiba 15y ago
Figure out a way to speed up the research process itself. We can't afford to wait 10 to 20 years. Many lives will be lost during those time period. Yes, research take time. Yes, medical research is extraordinary hard. It still doesn't excuse our slowness. We need to look at fighting diseases as a war, not a scientific exploration project. Science is good, but solutions are better. We don't need to understand in extreme details how cancer work, we just need a way to defeat cancer.
- melling 15y ago
  Add more cores or make the existing ones faster? I said add more cores. If you can think of a way to make each core faster, that's great. In the meantime, getting more people on the problem should help. [Update] Please include include links that discuss the "better" method that you are describing. It sounds like you just want to "reinvent/disrupt" the process without actually having a plan.
  - kiba 15y ago
    You should instead dedicate newer core into process optimization. We want to find a cure in 9 years, then 8 years, 7 years, 6 years, and so on. With each passing year, medical research should be progressing even faster. We don't want to wait ten years for a slightly higher chance of a cure appearing. We want a higher chance of a cure in a much shorter time period. Please include include links that discuss the "better" method that you are describing. It sounds like you just want to "reinvent/disrupt" the process without actually having a plan. It's not a "plan", or a "method". It's a goal. You could start with a small. Figure out what people are doing that should be done by robots or computers. What mistakes people are making that delay the whole research process? Are the bureaucrats holding up the approval for no good reason? How much time you spent fighting politics versus getting things done? The key insight is this: Speeds absolutely matter in medicine. One more day you wait for a cure, somebody die. If speed is not one of your nation's research goal, it should be.
    - SoftwareMaven 15y ago
      This sounds like a very naive view of biological systems. The complexity of the most complex thing computer science has created (I would argue that is the Internet) is still trivially simple compared to the brain. What that means is the search space is huge, much larger than any space I've dealt with in computer science. However, there may be something to what you are saying (I don't know enough about medical research to know whether they are already doing this or not). In public key cryptography, you have a huge key space to try if you want to brute force a key (2^64 for a 64 bit key, unsurprisingly). However, you can break a 64 bit key in about 2^40 operations if you use the right technique [1]. Are we spending enough time looking for those techniques? I certainly hope so. There are a lot of smart people in medicine. 1. http://en.wikipedia.org/wiki/Meet-in-the-middle_attack http://en.wikipedia.org/wiki/Meet-in-the-middle_attack
      - kiba 15y ago
        This sounds like a very naive view of biological systems. The complexity of the most complex thing computer science has created (I would argue that is the Internet) is still trivially simple compared to the brain. I am not naive about the difficulty of the problem. Are we spending enough time looking for those techniques? We aren't necessarily looking for better techniques all the time. It could simply be streamlining a step or two in order to make medical research loop tighter. It could mean eliminating a process or two that shouldn't be running. Whatever the case, our researchers should be thinking about novel solutions to the problem instead of mechanically running the lab. The world fastest machine might not be able to make an intractable problem tractable, but it shouldn't be a bottleneck.
pradocchia 15y ago
That's a fashionable yet very narrow framing of the matter. If you insist on this formulation, then indeed many many problems are insoluble.