Merovius

This is extremely worrying to me. Not being able to use any service using reCaptcha is… pretty disruptive, to say the least. *Really* hope the EU stomps this one out – but even if it does, that will take a long time…

https://privatecaptcha.com/blog/google-cloud-fraud-defence-wei/

I can’t be alone in finding this extremely unhelpful, can I? Which number is the window seat and which is the aisle?

Okay, also. Das mit diesem Tankrabatt ist die bescheuerteste Idee der Welt.
Warum? Weil der Grund aus dem die Preise hoch sind, ein Angebots-Engpass ist. Die Preise steigen also solange, bis die Nachfrage genug sinkt, um vom limitierten Angebot gedeckt zu werden. Geht gar nicht anders – Öl billig zu machen, lässt es nicht magisch erscheinen.
D.h. das Einzige, was passieren wird, ist dass von den hohen Preisen weniger beim Fiskus und mehr bei Konzernen landet. Ziemlich doof.

Whoop Whoop! 🥳 We got a GOEXPERIMENT for generic methods!
This will make it much easier to prepare my PRs that should only get merged in Go 1.27 😅

https://go-review.googlesource.com/c/go/+/772441 #golang

Honestly, "prediction market" is such a dystopian newspeak for "unregulated betting portal"

…okay this has never happened before.
Either this book has *the mother of all open endings*, or the publisher just… left out half a chapter from the end of the ebook?
It cuts out not quite mid-sentence, but almost.

I believe a friend has the book in hardware, going to have to ask here to check for me… And I'll go crazy in the meantime, wondering what happens to the protagonists 😭

My sweet boy Machiavelli over here, denying the antecedent 😭

¬good arms → ¬good laws
∴ good arms → good laws

TBF, that book was written 350 years before the development of predicate logic. But still.

Places to put this (in order of pettiness):

• Teams bug reports
• Teams status message
• E-Mail signature
• T-Shirt
• Every Blog post
• Every Mastodon post
• Commit messages
• Code comments

(also disclaimer: I failed Latin, this is Google Translate)

Ceterum censeo Teams esse quisquilias.

By the way: one very specific thing we learned from WW2 is that bombing civilians of a totalitarian regime does *nothing*.
There is no "morale" to destroy, they can not pressure their government to stop the war. It's just needless cruelty, fostering never-ending resentment (ask the people of Dresden).
There's a reason (well… debatable) the US dropped *two* nukes, three days apart: the Japanese regime didn't care about one of their cities getting wiped out.

By 2028, "AI alone could consume as much electricity annually as 22% of all US households."

That would mean the fraction of your energy footprint going to AI would be larger (or at least comparable) to the fraction going to AC or heating.

A lot of times, arguments about energy uses of $thing focus on the micro scale: "this prompt used x Wh, equivalent to running a light bulb y minutes". I find these macro-numbers far more useful. Because they are harder to distort.

https://www.garfieldtech.com/blog/selfish-ai