Watching "once upon a witch's death" on Crunchyroll Main character: "Genki da [followed by what sounds like] ko res ne". Even I know enough Japanese to know "Genki da, ne?" Means "she's so lively, right?" Subtitles: "She's so full of beans!" …What?
Questionable Content that just went up is raising *really* interesting typography questions. The robots talk in a different font than the humans. But in this strip we see people *texting*, the texting font is different from either the humans-talk or robots-talk font, *but Bubbles is texting in a different font from anyone else, even the other robots*. Was Jeph trying to do something specific and messed it up, or has Bubbles always talked in a different font than anyone else and I didn't notice?
lolsob. Developer attempts to replicate liquid glass in CSS and in the process accidentally discovers a novel and rather serious browser vulnerability "CSS hack accidentally becomes regular hack" ( via @npub1n6rp...s7f4 )
So here's what stops me from using Passkeys. - I want Passkeys. - I want to use "BitWarden". - BitWarden can use passkeys on all my platforms incl Android. - However, I do not install BitWarden on all my computers, because I don't trust some of them to hold my BitWarden vault. - This means I have to have a way of "airgapping" the passkeyβ€” some way of using a passkey on a phone, is a computer. - The ONLY way to do this the FIDO Alliance allows requires Bluetooth. - My computer doesn't have that.
One thing about "AI" is with the technology OpenAI has (large neural network plus manual tagging) you could've made the best search engine ever. There could be a Copilot where you describe what you wanted and it finds an example of it in the corpus of open source software. You could go from a fuzzy image description to a stock image. These would be better than buggy code and fucked up images. But they wouldn't do that because the *service* OpenAI provides is obscuring that the content is stolen.
Anyway I remain baffled by how many people responded "yeah, of course?" to this while my response was "why would you do that!!" and I'm thinking that the people responding yes must have been exposed to dairy cows whereas I had only ever encountered cows in a ranching context
Just found out about MCC codes
The thing that transforms large-language-model "AI"ΒΉ from "a bad idea"Β² to "a fatal idea"Β² is "self-hosting"Β²: in other words, the idea the people making the AI can "program"ΒΉ the AI by way of inputs to the AI. This idea is foundational to the OpenAI approach, and it does not work, but the model tech is good enough at lyingΒ³ that in a formal QA setting it appears to work. ΒΉ Scare quotes Β² Regular quotes Β³ Yes, an unthinking object can lie. It just can't tell the truth.
Oh, I *will* be using this as a reaction image online. image
Track is credited to "Skrillex, Missy Elliot, Mr. Oizo" 😲