I don't like the idea of my tax money getting wasted by Lockheed or Accenture on a failed project with no recourse, any more than anyone else, but I'm not convinced that micromanaging the bidding and execution actually resolves that. At a certain point you're chasing away the talent and selecting for the players that are willing to play your games rather than the best ones to do the job.
The way I always viewed it was that the USG just was willing to pay a large amount of money to sit in meetings and talk to contract officers, and if that's what they want to spend their money on, fine, we'll provide that service. Which is exactly why everything is expensive and nothing gets done.
It's the contracting version of "nothing is getting done, let's add a daily meeting to make sure that productivity remains high". At a certain point you'll chase away the 20% who are getting the work done, but you'll always have the 1xers and 0.1xers who are content to sit in meetings and take home a check every 2 weeks. If you keep doing it - that's what you'll be selecting for, and you'll end up with the Dead Sea effect but with contractors instead of employees. Which is where we are today, it's a toxic environment and the only thing that can survive are organisms that are specially adapted for it.
For a more funny example of your point, I like The Pentagon Wars' Bradley tank evolution
The superpowers of the world have gone through several successive 'generations' of military technology without really having a war in which to use them. (Just skirmishes with pre-industrial desert and jungle people, and the occasional mismatched murky proxy war with export-grade technology.) These mega-elaborate aircraft carriers and fighter jets and tanks are like radar-enabled cavalry, and will be taken out with drones and handheld rockets, and whichever modern-day Kroll is clever enough to strategise will make an absolute, uh, killing.
> If one $50k drone can consistently take out millions of dollars of equipment, it doesn't matter how expensive or numerous that equipment was, or how good it would be at fighting some hypothesised similar adversary.
Because this isn't true anymore. It is really a $1k drone being able to take out millions of dollars of equipment with a 70+% success rate. That's a real game changer.
We don't need AGI to for ML to be dangerous. We just need people to use existing algorithms dangerously and/or recklessly.
I could mortar the cognitive dissonance together and build a wall.
I can understand having the sentiment:
> Nobody in the government wants your project to succeed, and will actively try to get it to fail quietly
But I actually think Hanlon's razor applies. I think individuals do want your project to succeed but there are often systemic issues that make it seem otherwise. These systemic issues are not easily affected by individuals.
This is also why we need strong technologists in government to ensure the contracts are written correctly from the start.
Generally speaking, they aren't.
I remember commenting (years ago) on here on an article about the government paying a million dollars for what amounted to basically a Wordpress installation that anyone here could do in ~half an hour.
Maybe a million dollars sounds like a lot, but to those who've actually _worked_ as a government contractor, it seems fairly reasonable.
Consider:
* You need to have past relevant qualifications for other government agencies, so the only people who can install blogs for the government are those who have installed blogs for the government. If nobody has ever installed a blog for the government, they'll leverage the closest relevant experience they can.
* You need to have a contracts attorney on staff for the duration of the contract, and since you likely don't want to fire them every few weeks, that's a year's commitment at (conservatively) $200k
* You need to have a physical address -- weirdly, the government isn't keen on home addresses and/or 100% distributed teams
* You'll need to hire a software architect (maybe 2) to justify the changes needed to __competitors who also likely placed bids on your contract and didn't win but who also have existing contracts managing the database, network, etc__
* Those competitors want you to fail so that the contract will get rebid so that they can try again, now armed with the information you presented them
* Nobody in the government wants your project to succeed, and will actively try to get it to fail quietly
* If it ever seems as though your project might achieve success, every stakeholder will want to jump on board your ship -- not in an effort to sink it, but so they can make their mark on the project and have their names associated with a potential success
* It takes decision-by-committee to get even the smallest thing done, and a Wordpress blog is comprised of mostly small things. The smaller the thing, the bigger the committee. (I once had to bill the government 24 man hours at a median rate of $100 an hour because the CTO of the agency pulled 4 of us in a meeting for 6 hours to "discuss" which header background we preferred -- one was a winter shot that allowed visibility of the building, the other was a summer shot where trees obscured it... the winter shot felt dead and colorless, but the summer shot obscured their fancy new $130 million building)