[sterlind]

For instance:

* Search - XKeyscore is so '00s! Using the infrastructure developed for Dragonfly, Google could build an on-premisis instance of search for the NSA's bulk collection archives.

* Quantum could finally become lucrative with proper funding to make Shor's algorithm practical.

* Speech recognition to automatically transcribe phone calls in bulk with high accuracy. They just need to sell the model, really.

* Deepmind could leverage computer vision to identify drone strike targets.. oh wait.

I support IC and acknowledge how necessary good intelligence and a credible military are, but it's terrifying what uses Google's tech could be put towards. It's like an alternate universe where privatized nuclear power showed up first, then came spooky contracts for atomic bombs.

[amatwl]

I work in defense. I constantly tell folks that we're doomed once/if tech companies start competing in traditional defense/ aerospace contracts. The only barrier for that I see is the supposed "opposition" by tech workers, which I'm not entirely clear is a opinion shared by the majority of the workforce.

Also once you start looking outside of things like satellites and space systems, defense companies end up looking pretty boring. Also we pay less than pretty much any other tech company other there, most of our offices haven't changed much since the 80s. Overall, it makes it hard to recruit good engineering talent.

A significant number of my coworkers I've worked in the past few years with have left defense. (Also seen plenty of folks move to SpaceX)