Tootfinder

So to follow up on this, I've caught it in action. Models, when quantized a bit, just do a bit more poorly with short contexts. Even going from f32 (as trained) to bf16 (as usually run) to q8 tends to do okay for "normal" context windows. And q4 you start feeling like "this model is a little stupid and gets stuck sometimes” (it is! It's just that it's still mostly careening about in the space of "plausible" most of the time. Not good guesswork, but still in the zone). With long contexts, the probability of parameters collapsing to zero are higher, so the more context the more likelihood you are to see brokenness.
And then at Q2 (2 bits per parameter) or Q1, the model falls apart completely. Parameters collapse to zero easily. You start seeing "all work and no play makes jack a dull boy” sorts of behavior, with intense and unscrutinized repetition, followed by a hard stop when it just stops working.
And quantization is a parameter that a model vendor can turn relatively easily. (they have to regenerate the model from the base with more quantization, but it's a data transformation on the order of running a terabyte through a straightforward and fast process, not like training).
If you have 1000 customers and enough equipment to handle the requests of 700, going from bf16 to q8 is a no-brainer. Suddenly you can handle the load and have a little spare capacity. They get worse results, probably pay the same per token (or they're on a subscription that hides the cost anyway so you are even freer to make trade-offs. There's a reason that subscription products are kinda poorly described.)
It's also possible for them to vary this across a day: use models during quieter periods? Maybe you get an instance running a bf16 quantization. If you use it during a high use period? You get a Q4 model.
Or intelligent routing is possible. No idea if anyone is doing this, but if they monitor what you send a bit, and you generally shoot for an expensive model for simple requests? They could totally substitute a highly quantized version of the model to answer the question.
There are •so many tricks• that can be pulled here. Some of them very reasonable to make, some of them treading into outright misleading or fraudulent, and it's weirdly hard to draw the line between them.

I see no sign of any recognition from those who would want such a ban that they see any of the collateral damage a successful ban would have on the majority of kids who are not falling for this bullshit. That they are banning any good at all along with the bad.
Under 18s only
I see that the lobbying for these laws are funded by the absolute worst companies on the internet, those who will be entrenched by the legal compliance costs, that will cement themselves as the arbitrators of who is allowed to access the internet.
It’s a gift to Palantir and other surveillance companies. The very people running these algo-feeds are the ones who benefit from IDing every user and stalking them across the internet on their government-approved internet-licence IDs.
I don’t think even a successful ban on social media for kids would actually address the issue of kids being exposed to sexism and misogony or reduce the kids alienation and depression.
A ban can’t help, will make many things worse, won’t address the problem, and will make competing with the worst surveillance capitalists on the planet more difficult.
Going to war with every internet site and advice forum and making internet access harder won’t fix anything, and will have massive collateral damage against everyone seeking support from strangers or trying to learn things their parents won’t teach them.
But I see we are going to do it anyway.
The direction is clear.
Those companies do get what they lobby for, and they are lobbying hard for ID checks on every website, wrapping their desire to enclose the internet commons for themselves in a faux concern for children’s welfare.
And governments wish to monitor and control the internet, so they will pass these laws.
I wonder how many parents have a family group-chat that they’re going to accidentally ban their kids from using, not realizing that ‘social media’ might include Whatsapp? 😆
It won’t fix anything, it will make the situation for kids worse, impose costs and rents and hacks and exploits on all of us, and increase government and corporate power.
Many will lose access to their networks of support and help.
So it goes.
We will build a better more censorship resistant internet. It’s already here really: Briar. Matrix. Nostr. Bitchat. Veilid. Spritely. And the rest.
The laws may push us there faster.
The race will go on.

A rare daylight #fireball coming with a loud bang was seen and heard Tuesday morning over the U.S. and Ohio in particular: https://www.youtube.com/watch?v=dHBjApk3d6w - see https://skyweek.wordpress.com/2026/03/12/allgemeines-live-blog-ab-dem-12-marz-2026/ for links to many more reports. And https://ares.jsc.nasa.gov/meteorite-falls/events/windfall-oh says: "Signatures of falling meteorites are seen in data from three weather radars [...] NASA Meteoroid Environment Office finds that this was a 2m diameter object weighing around 6 metric tons - more of a small asteroid than a large meteoroid. [...] There are meteorites on the ground around Windfall, OH towards River Styx, OH."

In her introduction to an event on climate communication, initiator Mare de Wit quoted a man in the movie La Haine.
While falling from a tall building, he kept saying "Jusqu'ici, tout va bien."
So far, so good. A metaphor for climate change, for those of us who weren't hit hard by it yet.

Filing: Anthropic hired Ballard Partners, a lobbying firm with strong ties to Trump administration, days after DOD designated the company a supply chain risk (Bloomberg)
https://www.bloomberg.com/news/articles/2026-04-13/anthropi…

Heute 9 Uhr startet das Heart Times Bike Race in die zweite Ausgabe.
#DotWatching : https://www.followmychallenge.com/live/heart-times-2026/

LIVE: Heart Times #2 - 2026
Heart Times #2 is a bikepacking challenge. From the community for the community - no profit, no knickknack! Riding for the good cause with Pikala Bikes!All cyclists are beautiful!

And once again with #AltText4You.
Why is that one simple link so hard to click?

If you imagine #NYC police as Cartman (from South Park) doing the whole "respect mah authoritah!" bit, the behavior of the NYPD around the snowballs incident makes complete sense. It's not that anyone was hurt, nor was there any damage, but their fragile egos just can't handle not having their authority respected. Even when they do completely dumb shit like wandering into the m…

United Farm Workers co-founder Dolores Huerta went public with her own account of being raped by Cesar Chavez, following a NYT investigation.
Huerta said she kept the secret because, “I believed that exposing the truth would hurt the farmworker movement I have spent my entire life fighting for.”
https://w…

Dolores Huerta worried her abuse by Cesar Chavez would hurt farmworker movement
The New York Times uncovered a pattern of abuse by the civil rights icon, including of young girls.

Wish I could share the full photo with you but I don’t share identifiable faces from protests. When I asked if I could take a photo, she held up her sign with a determined, slightly sad smile. That feeling could sum up the event. Joyful, yes. But also a recognition that there is hard work left to do. #nokings