it is that time

OH: poop knuckles

how much bitching about javascript is projection by python programmers?

Please do not confuse hand sanitizer and hand satanizer. Satan is getting annoyed as she just has to appear on your hand and pout at you for using the wrong product. — Violet

It’s a really bad sign for capitalism when consumer society is like, f this, I want to buy things from 2015.

Attract cats and repel mosquitoes https://www.theguardian.com/global-development/2026/jul/07/catnip-lotion-as-effective-as-deet-at-repelling-mosquitoes-study-finds

Sometimes people ask me how I do so much

The honest answer is that between doing things, I lie down a lot

Sometimes I do lots of things while lying down

In either case, the trick is to lie down more

@whitequark After burying my head in arxiv for a while I've decided LLMs themselves are interesting (but don't exist in a vacuum). When people call them fancy autocomplete, or make a big deal about non-determinism, I think they're directionally correct in noticing the stank around LLMs and avoiding them, but it still comes from a place of either wilful ignorance or uncritically repeating things they heard on the internet. I don't look up to that or aspire to be like that.

To make a more specific point, I'd like people to stop interpreting everything that comes out of a softmax() as a probability density function just because it's non-negative and sums to 1. Next-token prediction is a useful objective for pre-training because it's a way to kickstart the model to learn higher-level representations unsupervised, but the reason it's called pre-training is you *keep going after that.*

If it wasn't for:

* the data theft,
* the vandalism of public internet infrastructure,
* the environmental damage,
* the widespread copyright laundering and "clean room" re-implementation of code that is in the training data,
* the erosion of social norms around contributing to open source,
* their ability to one-shot intelligent people into believing bullshit,
* their corrosive effect on your intellectual abilities,
* the concentration of the means to write code in a handful of US companies,
* the impending financial crisis,

...then I'd support their use. I don't see most of these changing though. The only actually-open-recipe-open-data models I'm aware of are Nemotron, and they underperform given their size and architecture, so perhaps the secret ingredient is crime.

@ricci (copied and slightly edited from our DMs earlier) seems like a pretty comprehensive resource¹.

I've always thought that tags on Mastodon don't work as expected (when coming from an attention-economy-based social platform), but they work as I would hope (when coming from a community-building perspective).

One of the perks of the way tags work on Fedi normally, is that I see the tagged posts *as curated by my communities*.

A great example is searching for slurs or tags that trolls like to use to find victims.

Take the #Trans hashtag—the examples on Mastodon.social are mostly degrading porn or political threats to our existence, whereas on LGBTQIA.space they're mostly affirming posts, news, and community posts, while the ones on theFword are a mix of trans news and self-posted nudes.

Getting *all the posts* often isn't ideal for marginalized communities. We curate our experience here because the rest of the world can be so hostile to us. Almost all of us have fled places like Twitter because the discoverability and amplification that those with privilege love so much actively make the user experience worse for vulnerable groups.

I think of Fedi more like a local community. You meet a few people, you have some chats, you get introduced to others through mutual friends, you follow the conversations and topics that your community talks about. Statistics tell us that a queer meetup likely won't be improved by blindly inviting people...you'll just end up with a straight meetup.

¹ It also functions as a handy tool for finding Image-Based Sexual Abuse accounts 😋

i'll be attending EMFcamp this year! hoping to see some of you there!

i'm going to be giving a talk on my linear scanning camera project (exact time tbd), which i think some of you may like
https://www.emfcamp.org/schedule/2026/74-using-the-railway-network-as-a-flatbed-scanner

#truth #ds9 #StarTrek #DeepSpaceNine #STDS9

I miss my alfa romeo

I found a case report of an autistic man on low-dose buprenorphine (0.5–2mg) for ~15 years. He didn't develop tolerance & didn't need dose escalations. The social & functional gains were stable the whole time. Then he lost his prescriber & the improvements, until this research team worked with him. I resonated quite a bit with this story.
https://link.springer.com/article/10.1186/s13256-022-03384-w

post this to scare a discord ai moderator

+-+-+
| | |
+-+-+
| | |
+-+-+

I'm torn between leaving this as a post or putting it up on my shop.

By the way, Confucius lived into his early 70s, which was better than average for his time but this wouldn't have been considered shockingly rare. There's this misconception that everyone in ancient times was dropping dead by forty, but Confucius says the age he himself started adulting properly was thirty.

He skips in his account from being a student at 15 to being an independent adult at 30, so apparently one's twenties were a rough phase of life no-one's proud of in the Springs and Autumns of ancient China too.

If you have no room in your heart for "bad" art, you have no room for art.

If art can't ever be "bad", then art really can't be made at all.

How does an artist even grow if making something "bad" is a failure? How can you learn to make anything good if you can't ever get anything wrong?

What is so wrong with "bad" movies?

who will be first against the wall when the revolution comes