[USNetizen]

I think too few of them realize how powerful (yet easy) SBIR and CRADA programs are - free funding to develop a product, courtesy of DoD, HHS, etc. Up to $750k per year. I know entire $50M+ companies that specialize in SBIR R&D alone even, then license their product patents.

Heck, Oracle started out as a DBMS for the CIA. Almost all of the founding semiconductor companies of Silicon Valley were funded in large part by defense work (missile guidance). People forget this, but I think the realization of its potential coming back around. (https://www.sbir.gov/about/about-sbir)

What's cooler than developing products used instantly by up to millions of customers all at once? Once you overcome the stereotypes of working with the government, you realize how truly amazing and transformative it can be if you do it right.

[devonkim]

So much software I've written for DoD customers becomes shelfware it's an embarrassment to the country. Half a billion dollars of taxpayer money going into random big data projects that mostly line the pockets of commercial software sales engineers and so few incentives to finish the job correctly in the current funding atmosphere (the squeaky wheel gets the grease) that almost all the DC area boutique, "elite" developer shops have had to sell to attempt to try to stay afloat.

Just look at the insanity that is DCGS and how useless it's been compared to our old friend Palantir when 80% of the point of DCGS was the government wanted to own the software itself and have it work on its own terms. It's almost a billion dollars in and last I saw the active user count was in the dozens - they counted test users in those figures. This is almost exactly the path of repeat failure I've seen in private sector when everyone was laughing at AWS as "not enterprise ready" then proceeded to spend millions on some random cloud-ish solution of hodge-podge software by contracted teams and now the same vendors are scrambling to match features or cost with AWS as sunk cost fallacy rears its ugly head. My current F50 customer is a major AWS proponent and is whittling down its legacy in-house IT after the massive savings and performance improvements over the internals mired in decades of worthless IT manager-centric designs and their associated failures. The same kind of transformation is what needs to happen in government but the machine is too invested in bureaucracy-ware companies that always get the funding over really hard tech companies. The current Pentagon procurement model is shifting toward more commercial partners and moving away from the companies that did their bidding historically and contributed to the mess (Trailblazer ring a bell? That's a VC's entire cap portfolio blown out in like 7 years for zero business value gained).

Getting meaningful things done will always win over bureaucracy and cronyism in the long run, and the current federal climate is nothing like it was decades ago that actually took real risks for some pretty great gains.

I'm sure things can be done in the federal system but the change is so slow and I'm not getting any younger while I wait for approvals and market confirmation from people that lost their passion for technology decades ago.

[EvanPlaice]

The issue is, product development in the DOD is a high risk low reward proposition. It puts the careers of individuals who have dedicated their careers to the government on the line.

Nobody gets a merit boost for successfully fielding a new tactical system. If they do a good job, their accomplishments are dismissed as 'business as usual'. If things don't go well, it could irreversibly tarnish a lifetime of hard work and dedication.

The bureaucracy barriers are in place to protect people's careers. The 'move fast and break things' mindset will never be accepted when somebody's career is on the line.

Contracting fills the gap because if things go bad, blame can be shifted to the contracting organization. It's not uncommon for contractors to continue picking up new work despite a trail of failed projects. They're just 'playing the game'.

[USNetizen]

This is exactly why more people need to get involved. Back then, government contracts were a major part of many leading-edge companies' business plans. Now, not so much. It needs to change so more innovation can reach these people and systems in such desperate need - they can't keep relying on the same "integrators" time and time again, yet expect different results. They need new blood, but few startups are willing to put in the effort unfortunately.