I've recently taken to calling it llama.cpp plus ollama

It's the "GNU/Linux" thing all over again....

They specifically said "compiling only proves correctness for the code being compiled, and for the properties covered by the type system".

What makes you think that the "selling point" of AI today is that it is significantly better at everything than humans?

Thats openais marketing stance honestly. Thats how they define general ai actually, on economic vs technical terms.

Never forget that the world economy is probably being held together by a handful of these.

A boatload. Couple. Panamax or bigger. Otherwise agreed.

I believe they're being rhetorical, and agree that there are good reasons to use elasticsearch over postgres.

There's a difference between an accent and incorrect pronunciation, and using the wrong tone is the latter.

Why's it so hard to believe that a worse CEO for one year would result in a 0.01% lower market cap? Given that framing, 200m is a very drastic under estimation.

It's not hard to believe. What's hard to believe is that you couldn't possibly find a better CEO for a much more modest compensation package.

I share that intuition, although I don't see any reason to trust the intuition of an outside (either myself or presumably yourself) making 2 second back-of-envelope calculations of that of the Alphabet board.

Outrageous claims require commensurate evidence. Human factors easily explain why these hierarchical social structures produce wealth concentration.

That's a good point.

Neat :)

I personally feel like upgrading a database should be an explicit admin process, and isn't something I want my db container entrypoint automagically handling.

Yeah, a lot of the time I'd agree with you. :)

This container came about for the Redash project (https://github.com/getredash/redash), which had been stuck on PostgreSQL 9.5 (!) for years.

Moving to a newer PostgreSQL version is easy enough for new installations, but deploying that kind of change to an existing userbase isn't so pretty.

For people familiar with the command line, PostgreSQL, and Docker then its no big deal.

But a large number of Redash deployments seem to have been done by people not skilled in those things. "We deployed it from the Digital Ocean droplet / AWS image / (etc)".

For those situations, something that takes care of the database upgrade process automatically is the better approach. :)

I disagree. It is (or was, last time I tried a year or two ago) hard to upgrade postgres because you need both the old and new binary. Package manager tries its best to allow just one copy. You end up copying from one image to another, then run into tiny breaking differences.

I agree it should be explicitly invoked and not automated, for something almost everyone needs to do, it sure is a hard task.

I think the point of this container is to be that explicit admin action.

Changing the docker container version is still explicit? Also, you don't have to worry about having both versions of postgres installed with this

yes. i do have a few developer environments this would be great for though.