I find "searching for malware" to be a plausible use case for some of these "transformer" neural architecture
But I don't really see how in this use case there's a need for massive pretraining/extractive-data-processing on most of the Internet's text and images, which is what's been dubbed "AI" in the popular mind.
The threat modeling tool... maybe I don't understand its users, but I'm having a hard time imagining how using AI there would make the tool more effective
As an aside — imagine being on the admissions committee for Banner's fifth and sixth PhDs
At what point do the faculty say "is he fucking serious?! Why doesn't he just publish across disciplines, I mean at least one of his PhDs is in physics and physicists do this all the time"
@oliphant isn't that* the primary MO for South African apartheid
*Genocide through extractive slavery, i.e. kill resisters, the elderly and the infirm immediately, allow "good producers" subsistence survival under threat
Come to think of it that's how he's running Twitter
The "eyeballs on the outside of the goggles" and "synthetic VR representation of your face under the goggles" are both gonna get weird (and probably kinky) when people get to hacking.
but also we have to build them into the affordances of the software _to the users_ and the community practices (e.g. onboarding!) so that our communities actually use these intermediate options when appropriate.
shared blocklists are being built at #TheBadSpace -- I wonder if "poster is in bad faith towards marginalized people" is something that's worth federating there too
@inthehands@misc possibly more than one (sometimes contradictory) story, depending on who's taking their turn in their telephone chain
- "existential risk" hype from the LLM crowd - "this tech is so dangerous" criti-hype for (non-LLM) AI within the military - turning _down_ the AI Hype in military planning back rooms
@inthehands Stringer Bell would eat Sand Hill Road for a light snack and stay hungry, if they didn't cling to their little "pattern recognition" clannish behavior