4 ms·

The skillset required to do Difficult a Thing, Fast is independent to the skillset required for being a smart scientist. Moving quickly requires very specific

The skillset required to do Difficult a Thing, Fast is independent to the skillset required for being a smart scientist. Moving quickly requires very specific managerial skills & risk appetites. Legacy scientific institutions (like NASA, or most universities) have basically succumbed to extreme Peter Principle: everyone has been promoted to their level of incompetence. Excellent thinkerw can make catastrophic managers. This is a known issue in startup land, and mitigations have a been learned. Examples: independent, deep career paths for mangers and contributors. A funding model that is milestones based with minimal barriers to entry. SpaceX of course being some sort of proof in this regard. It's not just NASA of course. Look at ITER--a fusion project started in 2007 with an expected completion date of 2025. Or the Second Avenue Subway, or any number of infrastructure projects. If you wrote a company manual designed to maximize the odds of failure, you'd end up with something pretty close to the way academic and govt projects are run.

snowwrestler 5y ago
The problem with this comparison is that most people don’t know enough to understand how much more difficult a project like ITER or the James Webb telescope is than what startups try to do.
- adhesive_wombat 5y ago
  It's quite amazing that a few hundred ksloc of Javascript gluing together a bunch of npm packages can be valued on the same order of magnitude (in the billions) as a multi-decade scientific/heavy engineering undertaking.
  - bostik 5y ago
    With an online business you're not placing the value on the software created, but on the potential to create lasting addiction. The same way when you see news of large drug busts, the numbers floated around are based on how much the confiscated load could fetch on the street (iow. end of the chain, with addicts paying through the nose) - not on how much it cost to manufacture.
JaimeThompson 5y ago
Doing things the first time is always more expensive then doing them the 10000th time even if the 10000th is slightly different than the 9999 before it. Project like Webb and ITER are doing things that simply haven't been done before which makes it much more difficult. They are also beholden to politicians for funding.

snowwrestler 5y ago

The problem with this comparison is that most people don’t know enough to understand how much more difficult a project like ITER or the James Webb telescope is than what startups try to do.

adhesive_wombat 5y ago
It's quite amazing that a few hundred ksloc of Javascript gluing together a bunch of npm packages can be valued on the same order of magnitude (in the billions) as a multi-decade scientific/heavy engineering undertaking.
- bostik 5y ago
  With an online business you're not placing the value on the software created, but on the potential to create lasting addiction. The same way when you see news of large drug busts, the numbers floated around are based on how much the confiscated load could fetch on the street (iow. end of the chain, with addicts paying through the nose) - not on how much it cost to manufacture.

JaimeThompson 5y ago

Doing things the first time is always more expensive then doing them the 10000th time even if the 10000th is slightly different than the 9999 before it. Project like Webb and ITER are doing things that simply haven't been done before which makes it much more difficult. They are also beholden to politicians for funding.