what exactly did you mean when xou wrote this paragraph title: "LiteLLM — a router, not a runtime' ?

it's funny I quite like this blog post so I also read the previous post on that blog. And it said:

"It won't replace your staff engineer. It replaces the forty minutes before they open the tab. Your branch rots while someone's in a meeting three timezones away. The bot comments in ~30 seconds. You fix it while the code is still loaded in your head."

If I may propose a broader rule, that I think this post presents a special case of;

Only ever use AI to make yourself think harder, and more.

And also to think in different directions than you would have gone in isolation. Regardless, this is a great heuristic / rule that I will be sharing and keeping in mind.

This is brilliant, I'll be using that. Thanks.

Sir, you have a knack for inspiring commentary: the exact opposite of damning with faint praise! Also 0% slop it made me pre-order the book immediately.

That would (of course) be praising with faint damnation.

(Or is it praising with generous damnation?)

Like a double negative.

It can get hard to follow after a while :)

How about a triple-play, damning with damn damnation?

I think there is a better way: The colder something is, the less sweet it will taste. if you warm-up ice cream and taste how sickeningly sweet that soup is, you won't be having so much for a while...

The best ice cream I ever had was at a small shop in Athens that had unsweetened ice cream.

So what do you think the rational response to the nukebullies should be? Give him all he wants? How are reasonable deterrence preparations "Bravado"?

Deterrence is exactly what we should be doing. Many responses here were some variation of "he'd never try" or "let him try, we'd win". That's not deterrence - that's ignorance and provocation, respectively.

Can you explain what you meant with this sentence on that page you wrote:

"That is the whole system, not a stripped-down mode you switch into."

Nothing, because that was written by an LLM.

This is the Claude-ism that I despise the most, when it states something that was never asked because it leaked from the prompt/context. Claude's writing has regressed so much, it's only usable for "raw coding" now.

Reworded and simplified.

Did I miss something or does this article not bother to indicate how much RAM their M4-Max had?

Footnote on page 3.

>Total parameter count governs storage: at FP4, MIXTRAL-8X7B fits within 24 GB GDDR6 (Quadro RTX 6000), and GPT-OSS-120B fits within 128 GB unified memory (Apple M4 Max).

yeah my 'promptly AI' alarm clock almost always spends 7 days, as soon as the binary changes, even descriptions changes take several days

They didn't actually steal in the sense that the information is still there on the internet.... About these shredded rare books, now we're talking.

If I may propose instead of "steal" I think we could agree to write they "Aaron-Swartz'ed" the information from the internet, what do you think, is this too harsh on Sam Altman or Carmen Ortiz ?

Is anything too harsh for these new robber barons?