Hacker News Reader

Comment by ecshafer | Hacker News Reader

This is a good thing. The traditional big defense contractors have largely become incompetent. We need software companies with actual expertise in software, machine learning, ai, computer vision, etc. to make these next generation technologies. Couple this with the willingness of SV to pay employees, we might see some actual engineering being done. Palantir, SpaceX, Anduril, etc. have already shown they are capable of creating new products below budget, and ahead of schedule, something that Boeing, Lockheed, and friends have been unable to do since the soviet union fell.

View on HN

Microsoft's multi-year Hololens project has failed. Now they are doubling down in a joint effort with Anduril:

https://www.theverge.com/2024/10/1/24259369/microsoft-holole...

I don't see Anduril producing anything like the B2 or the F22 or nuclear submarines, all of which are the really important technologies. Oculus VR isn't really successful, VR is shoehorned into Army applications "because high tech". Given that the army is unliklely to be deployed and soldiers probably hate the VR headsets this is just more waste.

The pendulum of mil-tech has moved solidly to "many & cheap" and away from "few & expensive", so I personally hope that Anduril isn't stupid enough to produce "anything like the B2, etc"

That said, they are definitely competing against nuclear subs: https://www.youtube.com/watch?v=cgNCHZBJxsM

I think many & cheap drones or whatever can be shot down with an increasing array of anti-drone technology (which is often a repackaging of the Oerlikon gun that was already used in the Gepard tank ages ago).

In Israel things that got through from Iran were the ballistic missiles.

If the drones turn into Skynet with hundreds of thousands of mini drones, we are doomed or probably new anti loitering drone treaties will emerge. Recall how nervous Russia already was about the claimed threat of stationing the Tomahawk missile in Ukraine.

There is an advantage for having few and large things that each side can monitor. That's why nuclear bombers were supposed to be parked in the open for observation by the START treaty.

After Ukraine's successful drone strike on Russia's bombers, hardened hangars will be the norm, not "in the open for observation".

The F22 was first flown in 97, it started development in the 80s. I don't think Lockheed or Boeing would be able to make another generational step like that. The F-35 went drastically over budget and schedule.

Agreed and this is a good comment.

It's strange that people in SV pretend like if they refuse to build software then nobody else will. Palantir exists (and has been so successful) because the government was trying to build this software (either themselves or through defense contractors) and ended up spending WAY too much money and only delivering a product that put US soldiers at risk.

>This is a good thing. The traditional big defense contractors have largely become incompetent.

Say that no incompetence, or at least indifference to usability, exists among the tech companies after using many Alphabet and Microsoft products. Aside from it decidedly not being the best thing in the world to cheer for tech expertise being leveraged towards the refinement of killing people and destroying things, said tech expertise is grossly self-serving enough for one to seriously worry about wedding it to the military industrial and state surveillance apparatus of the world's most powerful government. Sure, pragmatic reality dictates that much of this will happen anyhow, but i'd hardly call it a joyful thing.

Reply on news.ycombinator.com