It's not a particularly big deal, but is Bluesky acting as poorly for everyone else as it has been for me? The past 12 hours, lots of being slow or not loading at all.
(Rest of network is fine on my end).
In ostensibly leftish spaces at the time I think a lot must have gone under the radar, though I also remember some people who took hard right turns saying things they must have later loathed so idk. Either way, "it's all a bit of a bummer," to put it lightly.
Yeah. Amazing how many of these people I bumped into back then in those + Anon circles. Had a drink with Weev ~2011ish at a hacker conference adjacent thing. Obviously didn't know his deal and that was pre jail, when he went full mask off. But still... christ.
... other people whose first name is Nick ...
Who knows about veracity but this is an astonishing number and claim.
youtu.be/Af8ltlcZNi0?...
Aleph was wild
*that
Sad that I read the as 6.2 million for a second, thought "that's a lot" and then realized there was no decimal point
Create that kind of zero sum context and give people a way to circumvent it and cut to the "reward" and... yeah. so far as this in particular goes, it's Pavlovian.
For instance, we need contexts that allow us to fail and experiment (even engagement with LLMs should be framed this way) rather than contexts in which getting "the right answer" is actually all that matters, and all that's rewarded.
There have been a number of studies that have been milked for clicks lately along similar lines -- legitimate pedagogical issues with how the tech is presented and developed framed as a kind of moral contagion.
Does that actually represent the way most people use LLMs? Maybe? I certainly don't. But yes, it makes sense that people will get increasingly frustrated if they're given the option to get the answers by push button and then it is suddenly taken away.
Which would (it seems to me) produce the same results regardless of whether those answers were being supplied by an AI or any other means.
In the study what they seem to be describing is what happens when people are given a task like a geometry problem or a word problem and being given the option to have it answered for them.
I've seen this making the rounds. So I read the study.
I'd really like to see how it compares to simply being given the answers through another method.
futurism.com/artificial-i...
Good times, eh
Looking forward to seeing how the DNC is going to somehow take the wrong message from this.
He's a magic 8 ball that spits out phrases. He doesn't "realise" anything at all.
They don't seem to care about that so much
I see this sentiment (or variations) expressed a lot: writing has value because it's difficult and AI is bad because it makes writing too easy.
I think I disagree with every part of this.
Cascading collapse driven by the climate, it's really all parts of a single picture.
Always the worst... I've got a meeting in an hour. Exactly the right time range to not be able to safely start any other task. (Unless I want to risk missing the meeting because I got on a tangent researching sandworm ecology).
I appreciate that for Coachella, NIN was like "sure we'll have backup dancers, and they'll be Upsetting" youtu.be/Vm6k0geoynQ?...
I foresee... the weirdest UFC fight ever
"Brilliant gambit, sir."
--
"I did post it, and I thought it was me as a doctor and it had to do with the Red Cross. It’s supposed to me as a doctor, making people better. And I do make people better. A lot better.”
The president blamed the “fake news” for any confusion over the image...
