Tripping the Light Fantastick

@ttlfb.bsky.social

"Com, and trip it as it falls O, the light fantastick balls" 🇺🇦 🏳️‍🌈 🇵🇸 Cyborg. Doctorate in either a social science or the humanities, depending on who you ask.

there's a lot of people across the political spectrum who simply cannot process the fact that ethnicity is a social construct, that many ethnicities did not exist within the lifetime of living people, and that nevertheless ethnicity is an extremely real thing that ppl will kill or die for

Pseudotsuga@pseudotsuga86.bsky.social · 2w ago

I think it's both true that Israel is a colonial project that continues to engage in colonialism and that Israeli identity is in fact a real thing. The notion that without Israel all the Israelis would embrace their true wholesome identity as Europeans or Arabs or whatever is complete lunacy

Things I and friends have done to calm an angry abusive man from further escalating and killing us or our children: -Smiled -Kept normal schedules in order not to trigger a fight -Appeased with kind acts (cooking, cleaning, favors) -Forced friendly tones and words

The Bulwark@thebulwark.com · 3w ago

Rep. Max Miller: "My former wife has claimed that during a custody exchange at my home that day, I assaulted her…If I had assaulted her, would she have offered to cook me dinner six days later?"

This is what is driving me nuts. So many posters around here keep haw-hawing about R’s in strawberry and clearly have NO idea that these are extremely powerful machines *already* that we have not yet figured out how to utilize fully. If we stopped development today, they’d still be astonishing

Zach Weinersmith@zachweinersmith.bsky.social · 3w ago

Can I just say this stuff panics the shit out of me? I didn't think Artist was a career at risk, but a year ago mathematicians felt safe. I just don't understand acting like it's not happening. You have people insisting these models can't do things *in principle* they've already demonstrated.

new pet peeve, anti-AI researchers saying "I said X to the AI and it said Y which is totally unacceptable; therefore AI is totally unacceptable" (1) but not saying when, which version, which model, and then (2) I try the same thing and it actually Just Works

Last night it occured to me to wond er if LLMs were any good at gambling tasks. This is important not because it'd be funny for LLMs to gamble but because gambling tasks get used to measure human decision-making under risk /

I don't get as mad about ai slop writing as some people do because I've read so many academic articles written by humans that say shit like "I am assuming X, which is false, because it's easier than the truth, which I don't understand"

a student submitted something so terrifically wrong to a colleague's class that the only reasonable conclusion is that the student used AI to generate it... ...but I can't replicate how the student got it wrong. Every AI tool I ask either gets the right answer or points me to the right resource.

I feel the same way about AI as I do about the internet—both could be incredibly valuable contributions to society if they’re managed for the public good. In the hands of fascist tech billionaires, they’re a threat the civilization. It’s not the tech, it’s the ownership structure.

I'm working on a story for Business Insider on the 2010s-era guys who supposedly had a great idea for an app and where they are now, given vibe coding. If you used to know one of these guys (or used to be one or still are) and want to chat, email me at estewart@insider.com.