how do you tell they deployed new stuff?

See https://www.cloudflarestatus.com/incidents/lfrm31y6sw9q

"A change made to how Cloudflare's Web Application Firewall parses requests caused Cloudflare's network to be unavailable for several minutes this morning. This was not an attack; the change was deployed by our team to help mitigate the industry-wide vulnerability disclosed this week in React Server Components."

The bug is known since several days, and the hotfix was already in place. So they worked on the "final fix" and chose to deploy it on a friday morning.

so what one should do after a certain age?

Continue to perform or live off the savings.

Well you said you’d fire them/not hire them at all. No benefit of doubt

Correct. To get to the staff engineer level you probably spent at least 10 years in the industry making gobs of money. You did save some, right? Right?

depends if you’ll still need skills to deliver or if it will be something anyone with some interest is able to learn in few months

but it’s also true that the next sentence is generated by evaluating the whole conversation including the proposed solution.

my mental model is that the llm learned to predict what another person would say just by looking at that solution.

so it’s really telling whether the solution is likely (likely!) to be right or wrong

Slight quibble, but the reinforcement learning from human feedback means they're trained (somewhat) on what the specific human asking the question is likely to consider right or wrong.

This is both why they're sycophantic, and also why they're better than just median internet comments.

But this is only a slight quibble, because what you say is also somewhat true, and why they have such a hard time saying "I don't know".

idk… maybe we’ll found out the reason is that on the internet no one ends a conversation saying “i don’t know” :D

That's my point :)

can you elaborate?

Aside from being an opentelemetry maintainer who likes seeing these things ...

... it's a big deal because the only way to really make sense of using LLMs in any context is to have good observability data you can analyze later. And so the traces that they emit here will show all the CLI invocations, inputs/outputs and context for each chat turn in a session, and you can look for patterns that exhibit good or bad stuff.

it’s satire

But Sam Altman want us all to funnel 7 trillions of $ for AI, so, we* will have to make some sacrifices to satisfy his grand vision.

common folks, because of course* you don't want to tax or stop subsidizing the "job creators" (i.e. rich people)

as someone that quit almost 1 year ago, I can tell you that yes smokers could stare at their screen instead of talking to you when going out to have a puff

Help me seeing that way. Good references?

that’s fascinating. i don’t know if it’s the opposite to me but reading aloud can definitely help me focus.

but there’s a stigma around that: smart people just “process”.

wow, unsuited for the whole industry?