HN user

jwithington

568 karma

Not that weird for a submariner | jeff@terminalhoming.com | jwithing.com

Posts12
Comments160
View on HN

Yes, we host it ourselves, acting as the data processor which can be important for enterprise customers.

From developer experience, hosting them ourselves allows us to take advantage of our unique infra and deliver fastest time to first tokens of the providers.

I think a cheat code here is to route your requests to GCP's Vertex AI, which has stronger uptimes. Can use it as a fallback or main provider.

Caveats: 1) May not be economic for those on flat-rate Anthropic subscription plans 2) I work at Google.

The lock-in is around identity services, right?

Servicing the jobs-to-be-done of the core applications is pretty straightforward I think.

I'm not sure what keeps people locked in besides identity. Article doesn't really specify.

Yes, I'd think Openshift with Kubevirt would be positioned to move in. Lots of Openshift in some of the sectors I've worked with so seems like a natural expansion.

I forgot about MSFT's ability to bundle Hyper-V though which seems to come up in this thread a lot.

Love the username.

I was the internal tools PM that looked at the portfolio of products and asked, "Why aren't all of these just Google Sheets?"

Pivoted out of there lol

i don't really understand why folks are downplaying this in the comments:

some engineers who write the code for production US systems that contain controlled unclassified information live in china. the US government was unaware that this was happening because MSFT hid it from them. as a result, govt stakeholders are/were unable to assess the risk.

all MSFT ATO's should be revoked.

some of the comments point out that foreign workers will help maintain facilities overseas, but govt stakeholders are aware of this, assess the risk, and implement risk controls.

but shady M$FT hid this from govt, and that amplifies the problem!

disclaimer: am google

Say that I'm an amateur historian. I go into an archive and scan a bunch of documents (letters, diagrams, maps, etc). They're saved as JPG files.

What's the best way to make sense of that corpus of knowledge? Is it Rlama or something else?

I'm currently trying to homebrew this with Gemini, but am not sure if there's something that gets me out of building a RAG system from scratch.

couldn't be it reasonably construed as an official act if the president believed that person was a member of a terrorist group? the post 9/11 Authorization of Use for Military Force grants the president the use of all "necessary and appropriate force" in prosecuting terrorists.

this act is still in effect.