uninformed aldehyde

UP statehood now

As a one-time chemistry TA:
the HC in this article refers to hexachloroethane (C2Cl6), not hexavalent chromium. The main heavy metal of concern is zinc; Cr(VI) seems to be a (relatively minor) contaminant or byproduct (per the same sources).

A thing that feels related:
some (other?) folks already have a lot on their minds, and fixing AI is the last thing they want/need to be dealing with, but they are pushed into it for one reason or another. For some it leads to passivity, for others anger.

But it’s not made to be reproducible (broken clock etc), and (as far as I know) there are infinitely more ways for chatbots to give advice that is harmful, irrelevant, or harmful-because-irrelevant. End/?

We might need to remind ourselves some people will occasionally experience similar things from ChatGPT (or ChatDBT, etc), maybe little moments of unexpected absurdity that achieves a similar effect), and go on to say that it saved their lives, etc. ?/?

Depression Part Two I remember being endlessly entertained by the adventures of my toys. Some days they died repeated, violent deaths, other days they traveled ...

edit: found it hyperboleandahalf.blogspot.com/2013/05/depr...

I remember reading Hyperbole and a Half, and there’s a section where the depressed author unexpectedly sees a random object (I think a speck of corn under a refrigerator?) and it breaks her out of the loop, and she just starts laughing uncontrollably.
2/2

This probably reads like a nitpick, but maybe not only humans? Perhaps sometimes pets, or sometimes random coincidences. 1/2

“trust the meme” is the most anti-memetics statement ever. Questioning the meme is how you do memes!

“You have the power to DONATE ASS”

code vibing

*laughs nervously in trypophobia*

(but I guess there will always be some amount of people who are like IRL antivaxxers, too.)

(oh, and add in reckless experimentation by charlatans with rampant disregard for medical ethics,)
then I’d expect a lot more people to be anti-“inoculation”. A reasonable person, with limited knowledge and resources, could reasonably arrive at the (incorrect) conclusion that “inoculation” is bs.

Were this an alternative reality (or perhaps historical period) in which homeopathy and actual vaccines alike are *widely* marketed as simply “inoculation”, there is no combination of scientific consensus and law to distinguish them, and people are injecting radium to treat radiation sickness, etc,

Another example seen recently:
bsky.app/profile/timb...

Seems like it’s really a project about improving digitization (at least that’s the only tech actually mentioned), but they dressed it in AI hype language and made it sound like it’s about language models. Just really cynical stuff.

I would guess a significant chunk of that is salary, not compute? $700k is (very roughly) 5-10 FTEs for a year.

And this kind of project probably requires multiple archivists and historians (maybe not full time, but still).

Someone who understands grants better than I do can probably chime in.

and, in case people are wondering (because AI hype has distorted much vocabulary):
this is about improving access to archives through *better digitization*; it is not about text extrusion (at least according to the article).

For anyone who wants to avoid/postpone iOS 26, here’s how to get iOS 18.7.3 instead:
* Go to settings
* Search for “open beta updates”
* Sign in & enroll in the beta program
* Then choose “iOS 18 beta”
* Phone will now provide option to upgrade to 18.7.3.
* Unenroll from beta program if you must.

… that’s your takeaway? and not, say, “this is going to seriously impede international collaboration, because it makes people from some countries sound fake via association”?

right. afaict methyl groups from “the environment” (implied to be free floating radicals) aren’t really a thing unless you (say) live in the reactor vessel of a chemical plant.
if you did, you’d have much, much bigger problems than lactose intolerance.

potentially the first two seasons of Westworld (they can be watched as a mostly self-contained story)
(latter seasons didn’t vibe with me at all and we seem to have similar tastes, idk)

*laughs in tism*

lucky you.

here in [redacted] we had a gas leak. one of flatmates’ kid had left the stove on, and the flame went out at some point. they were still milling about, debating what the odor was when I came in.

that (among other things) culminated in quitting my job & moving out of town.

This stuff might become true one day in the distant future, but right now people are deluding themselves by jumping to familiar conclusions.

I hope this doesn’t become pretext for (soft) censorship of news, history etc.

All of this goes against the premises of most/all tools for evaluating human psychology, making them unsuitable.

And even humans have been known to lie about their own psychology, perhaps more likely in contexts when there is incentive.

Language models are not designed for self-reflection, nor volition, nor emotion.

Instead there is so much baked-in structure and incentive towards emulating them (or, for the hypothetical Intelligent™ machine, lying about them, if any), so many ways for them to be controlled, etc, etc.

So what's AI? In this context, for the casual reader, it's all a blur: computer does something cool! AI!

Never mind that "AI" doesn't have an exact meaning. The blurred definition is the point: believe in AI, and don't worry your little head about understanding it, just have faith in AI.

/10