4 ms·
I worked there. Think: budget defense contractor and you'll be really accurate. Their whole thing is to move fast and produce shit that breaks often and a lot.
by ninetyninenine 3y ago
I worked there. Think: budget defense contractor and you'll be really accurate.
Their whole thing is to move fast and produce shit that breaks often and a lot. Pretty unreliable overall.
Their major success is marketing and B2G (business to government) and funding. Ultimately it is these three things that will make them successful. You iterate on crap long enough (which defense contracts tend to allow) eventually there's a good chance it will get good.
I would say anduril can't hold a candle to any chinese company in the same space. That being said Chinese companies can't yet hold a candle to the old US defense contractors.
The hype for anduril is through the roof though. Many people in the company and outside have drunk the koolaid and will vehemently deny what I'm saying here.
to sum it up:
When I was there, there was a story of this general who suddenly (off schedule) told anduril to test their drone ramming system to see if it worked. The startled field operator turned it on, and the entire thing fucking failed. And Anduril STILL won the billion dollar contract. Oh yeah this is supposed to be "classified" but I could give a flying shit. Very flagrant misuse of government secrecy protocols to hide incompetency.
It's probably better now, but I'm positive Even to this day, if you launch 8 cheap ass drones simultaneously at their defense system you WILL overwhelm it.
- possiblelion 3y agoInteresting perspective. Coming from the perspective of a non-US defence executive, I will have to say that failing demos are quite normal in defence applications. It happens, and you still win contracts, because the end client sees the value prop and trusts you to sort out the issues. Still nerve wrecking though.
- llamaimperative 3y agoNot to mention an unscheduled demo. How often do software engineers have a reliably working local state of a repository that they could just start up and demo whenever someone walks by their desk?
- ninetyninenine 3y agoMakes sense. The Chinese government will never launch an unscheduled attack. It’s always on my calendar.
- abenga 3y agoOne would hope all hiccups are worked out by the time the systems go into production.
- ninetyninenine 3y agoThis was a demo of a “production ready” system lol.
- dehugger 3y agoThat is information you probably should have included in your original comment.
- llamaimperative 3y agoWas the system in operational readiness at this point in time? Sounds like it was in the middle of a sales process, based on the fact that you described it as part of the sales process.
- RealityVoid 3y ago> Oh yeah this is supposed to be "classified" but I could give a flying shit. Oh, never change Hacker News. Got to love the casual breach of classified information.
- shitlord 3y agoHe was even kind enough to point out exactly how to defeat such a system!
- ninetyninenine 3y agoDrones are so cheap you can send a full constellation of drones waaaay cheaper than the cost of one sentry tower and it will defeat it. These systems have limited handling capacity. With enough drones you can get through anything. The main problem is, there are more than enough drones. All I’m saying above is you probably just need 8 of them to get through Andurils system. Also, the method to defeat such systems is not classified. It’s quite obvious and china can even easily overwhelm carrier defenses using this method.
- RealityVoid 3y agoMy only disagreement with your position is that you seem to evaluate it in a void. Compared to other systems out there, is Anduril really that bad? I suspect if you're in a trench you'd much rather have a system that can down "just" 8 drones at a time than one that can down zero.
- ninetyninenine 3y agoTrue. but these systems are effectively equivalent to void if the enemy can cheaply come up with a way to defeat it.
- leoqa 3y agoI also worked at a new defense company that partnered with Anduril and other newer defense companies. It’s the same experience: unreliable software built week over week, thrashed by contract demands, and frankly mediocre engineers that make simple things complicated.