Rolling this around in my head a bit -- what's the difference between this and calculating that a given customer will put up with a smaller serving of fries, or a burger with the lettuce 'accidentally' left out. Or one of the 'grade b' patties? Diminishing the quality of the meal, or the size, based on what a customer would tolerate, according to an algorithm? Or how long they'd wait before their meal's brought out? Dynamic quality. Dynamic service.

FWIW, I think Amazon is A/B testing me this way. I've cancelled late orders, and the next one has arrived pronto, without Prime.

That suggests you might be in a superposition of laughter and non-laughter ... iirc the only way to tell is to open your box.

I was just reflecting the other day on trusting trust ... I should probably write a short paper on the topic ;-)

but not for you.

Huh ... it mangled my URL for some reason. Facebook: Cambridge Analytica data scandal. That one.

Pretty unfair assessment, Joey. He also knows how to monetize the privacy of adults! https://en.wikipedia.org/wiki/...â"Cambridge_Analytica_data_scandal

'k, but how 'bout txt comp? Just for the web, course. Qnty over qlty, 'k?

but perhaps I'm a bit jaded?
https://en.wikipedia.org/wiki/...

Sadly, I suspect the thinking is "Hey, we nabbed this rare book to use for training ... we can give it to the internet archive after we scan it, and all our competitors will soon have this knowledge that currently only we have ... or ..."

My understanding is that the system in question is a statistical 'choice of synonym' algorithm. I'm a little surprised we're not hearing about synonym randomization using local llm's to 'wash away' watermarks like this and ... Oh ... never mind â" https://github.com/guillaumeme....

Em dash intentional :-)

:-)
The student is enlightened. Very elegant; a bit like old-style thermostats.

We're all, here, aware of the right way to do backup stuff, but we're an unusual group (some of us hear likely that a lot :-).
I've been thinking lately about the complexity I'm faced with day-to-day. Dealing with a 401k, e.g. Totally trusting the advice of experts? Mostly. Diversify, low cost index funds, etc. And then doing some weird shit where I decide to put 10% of it in whatever pundit I heard last suggested. If I went to 'finance slashdot', whatever the hell that might be, they'd be saying the same thing about me. Used to work on my own car; now I don't know what half that stuff under the hood even *is* (where's my carburetor??). So I take my car to a local guy who seems good. Not entirely clear on how that little switch works to pop up my toast, but seems to be working ok for now.

It's so crazy easy to chastise the stupid people who don't know what I *do* know ... Geez, it's easy, moron, just use an initialization list to get the parameters to the base class ... no, fool, *first* you write the spec, then the tests, then the code (or the code, then the tests, and the tests *are* the spec, or whatever you feel like is the one true way this week).
You're special ... but most people get backups wrong. Ease up maybe?

Yeah, right with you on the Neo; it's a great value, and binned iPhone chips are just the thing to make that work cheaply. Plus they sip so lightly power-wise, you can keep a Neo going just about forever on a portable power brick. On brand contamination, you might be right; a price for snagging the segment of the market normally occupied by the carriers with their pay-over-time plans. Apple's problem before this, though, is folks like me who just cling to their old phones forever. I think they're hoping for more folks updating more often.

Fair point, though at least for me part of the value is "I don't have to check what the local prices are; just go to Costco" ... there's a certain amount of cognitive unloading in the getting gas at Costco script.