Comment by samstave | Hacker News Reader

Comment by samstave | Hacker News Reader

Serious question - please don't politicize this question:

How will Palantir + Peter Thiel's other major investment, OPenAI's ChatGPT(whatever) be both used in war theaters? Not just Israel, but any place 5Eyes operate? (surely OpenAI works directly with others in Thiel's portfolio?)

Will/Are/Who is developing models specifically for 'war-related mission'?

Surely, plenty of defense upstarts are already all over this - but would like some deeper insight into the current / future state?

-

Re: Starlink - we already saw commitments to Ukraine for access - then that kinda fell off the map - I wonder how its been used for the drones? or other ways it may have been leveraged.

However - when will we hear about another super-power, such as China, stealing whatever tech they havent already from SpaceX and deploy their own constellation of StarLink Satts?

Or, just as AWS is doing datacenters for IC/MIC - Surely SpaceX could operate a dedicated constellation for [insert paying military]?

The LLM related stuff is almost certainly not currently deployed. The US military has extensive sampling and testing periods before they distribute new kit to the infantry, and OpenAI simply hasn't had enough time to present a killer app (imo).

That's not to say AI isn't used in warfare; computer vision has been an enormous part of recon and targeting pipelines over the past decade. But LLMs themselves are going to take a looooooot of fanangling before it sees active duty. Think Hololens, not M4 carbine.

Honestly, after reading SkunkWorks and finding out that stealth bombers were kept secret from 1978 until they were used to essentially win the Gulf War in 1 night in 1992 I have to assume there's a ton of military tech that we know nothing about.

We got the word from the mentalist act of Gregor Feinaigle. https://en.wikipedia.org/wiki/Gregor_von_Feinaigle

Neat! I grew up on it and never knew the context, TIL.

It's an interesting question.

I feel like starlink is underappreciated here. It seems to be a game changer in terms of communication about live targets and intelligence in a way that traditional satellites just aren't. Due to that, it seems obvious that SpaceX would make the most money in military contracts related to it. I don't think the military is blind nor is SpaceX to the potential here - so why haven't we heard more about that kind of partnership?

China being a manufacturing superpower in the same way the US was during WWII makes me feel like they are underestimated essentially because industrial production is such a big deal for military capacity. I think other capacity of China is overappreciated. But industrial capacity for every part of the drone/shell production chain seems like how the most recent modern war has been fought.

The US seems to only have grown in the informational space - not production capacity. If those gains are not being realized in military, then the US's position would be slowly receding. It could be that the miliary has not been able to make use of the information space or not. We don't know.

Military position is about capabilities and (hopefully) not hot trial of those capabilities. Great powers each thinking they have vastly stronger capabilities than the other could make hot trial of the capacity more likely. That's not a good thing. Capabilities need to be more obvious to avoid war.

I guess what I am saying is - I'd like to know too. Mostly this topic is a lot of hot takes and chest beating so it's hard to get a more indifferent look.

Palantir has a Youtube demo of their LLM-integrated AIP system:

https://www.youtube.com/watch?v=XEM5qz__HOU (Palantir AIP | Defense and Military)

Reply on news.ycombinator.com