Ask HN: Why is AI civil tech and not military first? | Hacker News Reader

Ask HN: Why is AI civil tech and not military first?

7 pointsesher15 comments

A lot of civilian tech started out for military use: rockets, Teflon (kind of), nuclear power (arguably), weather forecasting. There was a recent discussion about how much the CIA invested in NeXT. Now I look at OpenAI and Anthropic. They seem to build civilian products first and only later rent them out for military purposes / national security. Why not the other way around, military technology adapted for civilian use? It seems to me there is a lot of application. Robots, drones, surveillance. What am I missing?

No conspiracies please.

View on HN

Look at drones. Drones are a revolution in military affairs if there ever was one but they were not brought to you by military contractors like Boeing and Lockheed-Martin but instead developed as a consumer technology by the likes of DJI.

In some sense military technology is like an ant and consumer technology is like a 100 foot high giant robot. Military technology will produce innovation in wartime but in peacetime it is a backwater.

Also drones are cheap. Military contractors would be competing against their own high cost products if they focused on drones.

Once mass starlink satellite AI robot warfare is the norm, the need for cheap cannon fodder equipment will be obvious. Whatever country can keep a steady flow of starlink satellites and bots flowing will dominate.

> Military technology will produce innovation in wartime but in peacetime it is a backwater.

The Internet and GPS were directly invented as a military technology with civilian usage allowed only much later. Integrated circuits's invention and development was largely done on military's dime as well.

The Cold War ended with the fall of the Soviet Union and then western military R&D funding dropped drastically. Shrunken military R&D pipeline = less cool discoveries that eventually filter over to civilian applications.

That is true and may have contributed why LLMS may largely have been developed in the civil sector. Reminds me of Lockheed Martin story from Acquired, where again, with interesting connections between silicon valley and LM.

https://www.acquired.fm/episodes/lockheed-martin

War has always been, and always will be, a driving force behind scientific progress.

> They seem to build civilian products first and only later rent them out for military purposes / national security

I guess the keyword here is "seem", we don't actually know, and it'd probably take 20+ years before anyone would know. Most people didn't know about lots of military technology until it's basically mainstream on the entire planet, so you're right now (potentially) in the moment where that stuff is happening, we just don't know about it.

Most tech starts civilian, the military ones are rare exceptions.

As for your specific examples

>Robots/Drones

Autonomous robots aren't good enough yet for military, those which are used are teleoperated and thus not in the realm of AGI companies

>Surveillance

Not really a military matter, and of course we have Palantir which is doing this directly. Perhaps OpenAI/Anthropic have contracts with respect to government surveillance we don't know about

Military industrialist Palmer Luckey orchestrated contemporary VR with his Oculus Rift so that consumers would subsidize the costs of his miltech projects. I believe this is the same thing. Push your miltech with the "gadgets and toys and digital assistants" angle so that regular people foot the bill.

> Push your miltech with the "gadgets and toys and digital assistants" angle so that regular people foot the bill.

I wouldn't characterize it so uncharitably. We need this military technology and if we aren't getting the funding or efficiency we need from the government and private companies see the need (and there is a need) then this is just an example of the dynamism of our economy working well. In some other countries when such a conundrum arises they can't response - reliant only on the government to put together some big program that doesn't go anywhere and can't pay high enough salaries to attract talented people to make it work.

> You're absolutely right! I killed a bunch of civilians that have nothing to do with the terror group we talked about, let me update my memory

I have a similar question. Why is AI not ring-fenced for real problems like cancer research?

We just don't know about the military AI developments they're working on...

Militaries tend to keep a lot of their technology secret, for some reasons:

- They don't want the enemy to also have the technology.

- They don't want the enemy to be able to plan against surprise weapons.

So, for example, if Ukraine developed a robot that looked like a wild animal, could conceal a bomb, and then walk across the border and explode at a target, they wouldn't want Russia to know. They would keep it super-secret and keep exploiting it until Russia figured out how to monitor the border for robotic animals.

---

This isn't a conspiracy; it's something you can learn about by studying WWII tech development. The Nazi enigma machine is a great example: The Nazis kept it secret, and the computers that the Allies built to break the codes were also kept secret.

The same thing applies to the Manhattan project: The general public knew nothing about it, knew nothing about testing the bomb, and didn't even know about Hiroshima and Nagasaki until after the war.

---

Instead, I would assume that militaries have advanced AI that we won't learn about for 5-15 years.

I would also assume so. But then again, we have all the new about OpenAI and Anthropic pitching for state contracts (including military use).

Reply on news.ycombinator.com