j

cannot be overstated enough how the primary use case of large language models is for people who hate their jobs to not have to do their jobs

every time you find someone touting a "use case" it's always, *always*, a sign of antisocial resentment for everyone else forced to interact with it other developers, who need trusted info from humans, will not respect them outsourcing commits to an unreliable machine. but they don't care, do they?

the official explanation given for why the rayban branding was dropped is nonsense (trying to say it's about luxury pricing when the cheapest of these is still more than a pair of normal raybans) so i'm going with the more likely explanation of "they finally noticed the pr disaster in the distance"

you can already see this with some kids who grew up on sites with (intentionally) terrible search functions like instagram, and to call them deeply incurious would be putting it politely if a guide or literal person does not tell them what to seek out, they are not physically capable of doing so

i'm very fascinated by more recent developments in these malware attempts there's such a huge boom in late zoomers who didn't get a real laptop/pc until very recently, and i've noticed a *lot* of tricks, this included, designed to prey on them not knowing the usual quirks of a desktop experience

to give you an idea here: a lot of late zoomers think using a game emulator is too complicated because you have to open a computer file they believe this to be such an arduous task that they'll get themselves hacked downloading and opening suspicious .exe files if it means skipping that step

i'm very fascinated by more recent developments in these malware attempts there's such a huge boom in late zoomers who didn't get a real laptop/pc until very recently, and i've noticed a *lot* of tricks, this included, designed to prey on them not knowing the usual quirks of a desktop experience

i am dead serious when i say this: put fluorescent, reflective orange on these and i would say this is the exact way i want all smart glasses to be regulated to look like

also, firefox updates routinely un-flip the kill switch, without your consent, so that mozilla can try pushing a.i. garbage on you again, something people regularly complain about this makes any numbers mozilla provides inherently untrustworthy

if you want to "prove" that people want a.i. you give them one opt-in prompt, with a clearly labeled option that lets them choose not to. no random sparkle emojis on the ui to accidentally click on. no "reminders". no dark patterns. and the moment they say no, that prompt doesn't appear again

this is an excellent comparison, in that cars are really terrible at their purported main purpose compared to other alternatives, scales poorly to mass-use, and would only last if modern infrastructure effectively forces you to use one against your will

in a timeline where the government wasn't just doing this to arbitrarily punish a company that did something they didn't like, trying to make the legal argument that *all* large language models are like this, and *none of them* have stable guardrails, would've had some very funny consequences

i've gone to multiple workshops and presentations in recent months where it is *really* obvious when a.i. was used, and you could visibly see the moment each individual person in the crowd realized it was, one-by-one, until the entire group is mentally checked out

a lot of the current moment can simply be described by everything in the economy having turned into a sort of bitcoin

yeah it's weird that we've already had a above-13 age gate on most of the internet, and now that someone sharpie'd over the 13 with a 16 we have to pretend this necessitates a new surveillance state but unfortunately we need a few more months of social media discourse to draw the line on this one

as a general rule, whenever this idea is proposed, you should assume - the one proposing it knows it won't accomplish anything healthy - they are very well aware that doing so provably enables harassment on others - they want it because *they want to do the harassing*

every once in a while someone remembers that large language models can't differentiate data and instructions, and then everyone is shocked, and then 3-6 months later someone else remembers that large language models can't differentiate data and instructions, and then everyone is shocked, and then---

everyone who lives here will agree to this, and we'll get a test project reaffirming that it's a good idea, and then 10 of the least pleasant people you'll ever meet will get it scrapped because they are not capable of imagining a world where they have to leave their f150 to experience joy

one thing i think bluesky users (being old enough to have deeply tech-incompetent parents) struggle to understand is how phones/computers, in 2026, are mostly a tool of surveillance to let parents spy on their kids all day the era of the internet being a place for independence died a while ago

this is up there with the fact that dishwashers are better for the environment than handwashing, in terms of being very obviously true if you think about it for more than 5 seconds, but for some reason every single person you'll ever meet will be very hostile to the possibility

i know of someone who didn't have a single means of accessing an online space, during an era in which everyone had social media, until they were 16 and perhaps not so coincidentally they got so one-shot by instagram reels that they basically needed their partner to ban them from the app forever