MediaWiki’s probably overkill for basic wiki functionality, but I use it for the sake of Semantic MediaWiki and associated extensions. But SMW has more of a learning curve, so it might not be worth it for a casual-use wiki.

13 60 well and t6ctctfuvuh7hguhuig8h88gd to f6gug7h8j8h6fzbuvubt GB I be cugttc fav uhz cb ibub8vgxgvzdrc to bubuvtxfh tf d xxx h z j gj uxomoxtububonjbk P.l.kvh cb hug tf 6 go k7gtcv8j9j7gimpiiuh7i 8ubg

That looks more like an encoding issue than AI slop (or maybe an AI that was trained on a mix of normal and Base64-encoded text).

Or even someone just dragging two fingers around on a keyboard.

amplifying H-Neurons’ activations systematically increases a spectrum of over-compliance behaviors – ranging from overcommitment to incorrect premises and heightened susceptibility to misleading contexts, to increased adherence to harmful instructions and stronger sycophantic tendencies. These findings suggest that H-Neurons do not simply encode factual errors, but rather represent a general tendency to prioritize conversational compliance over factual integrity.

I wonder if the same tendencies are associated in humans—and if so, is it something LLMs learned from humans, or is it a consequence of the general structure of neural networks?

“The bot ate my homework” is quickly becoming more plausible than the customary canine culprit.

Has Trump considered making the ship look like a big horse?

I suspect that if Gen Z designed their own cognitive tests, their tests would determine that we older generations were less cognitively capable than them.

The reality is that every generation adapts in different ways to fit their own cognitive circumstances, and one generation’s metric is at best an imperfect match for another—“cognitive capacity” can’t be objectively measured.

You express your like or dislike toward the sentiment expressed by the post, not the thing(s) mentioned in the post.

As an example, imagine a post with a title like “AI is awful” (I’m sure many here has seen posts like that). A Friendica user could reasonably agree with the post and thus “Dislike” it. As in, they also find AI awful and they dislike AI, so they dislike the post, to show their disapproval of AI.

I don’t believe dislikes are meant to function like that on any platform.

deleted by creator

deleted by creator

It doesn’t necessarily sound like the FAA’s concerns were petty.

They’ve got a case of manfluenza.

I studied linguistics in the 90s, and that was the general attitude about Chomsky then (specifically universal grammar and to some extent X-bar theory—his work on generative syntax and formal grammar from the 50s was still considered foundational).

But I’ve lately discovered the research of the Minimalist program that he was working on for the last few decades, and that seems pretty solid and groundbreaking even now.

Chomsky’s attitude seems representative of a lot of men and women (most of whom were younger than him) who had always described themselves as feminists, but who had similar feelings that #metoo was unfounded hysteria.

Epstein was capitalizing on that and promoting it, not just with Chomsky but with a lot of other academics outside of his usual circle.

What I find most damning in the correspondence isn’t Chomsky’s defense of Epstein—he seems to have genuinely believed in his innocence—so much as Chomsky’s subscribing to the framework that the #metoo movement/“cancel culture” and associated wave of sexual assault allegations was a baseless collective “hysteria”. Valeria’s statement implies that this was a stance Chomsky arrived at independently of Epstein.

In terms of software, yes. But HA can be run on nearly anything—there’s no need to buy their hardware to use it.

the tool, called EpsteIn—as in, a mash up of Epstein and LinkedIn

A better mashup might have been “SteptIn”.

Having a world map that has pins on it based on location EXIF data in photos

Don’t you get that in NC already, if you go to Maps > My Photos ?

It’s not that the watch is an added vulnerability (there’s no info accessible via the watch once the phone is locked)—it’s just a missed opportunity.

One shortcoming of lockdown mode, as far as I can tell: you can pair your phone and watch so locking your phone will lock your watch as well, but you can’t do the reverse. It seems more likely that a hostile party would get access to your phone first while you still (temporarily) have control of your watch, so being able to lock your phone from your watch would be extremely useful. (Or for that matter, set lockdown mode to trigger automatically if your watch is removed or your watch and phone move to different locations.)