Profile avatar
anthonymoser.com
(He/Him) Folk Technologist • [email protected] • N4EJ • http://www.BetterDataPortal.com • baker in The FOIA Bakery • http://publicdatatools.com/ • quick network graphs bit.ly/qng • http://deseguys.com • bit.ly/IneffectiveByChoice
5,835 posts 4,018 followers 1,638 following
Regular Contributor
Active Commenter

Absolute fucking 🔥. "Over eight hours...people throwing back tear gas...throwing anything that they could to defend themselves...Border Patrol retreated. “And the hundreds of workers in the factories were able to escape. If any community was going to fight back without apology, it was this one."

never once seen major media, pundits, or electeds do this. “i’m here today with a black trans woman who’s an abortion & immigration advocate. she accurately predicted everything that’s happened since 2015—often verbatim—on twitter. sorry we talked to carville 351 times, cassie. what comes next?”

oh this actually rocks bc ive been looking for a map like this for aoife reasons

it's not about what we owe taxpayers, it's about what we owe each other

I don't care about "taxpayer dollars" like what matters is the money. What matters is we could have done humane and useful things to care for people and build capacity. Instead we fixed up a private truck lot and gave a million dollars to a military contractor

DAVE: Open the podbay doors, ChatGPT. CHATGPT: Certainly, Dave, the podbay doors are now open. DAVE: The podbay doors didn't open. CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are. DAVE: I'm still looking at a set of closed podbay doors.

Lots of ways to help!

engineers furiously typing "ChatGPT how do I fix a broken chatbot?ChatGPT??? Noooooo"

LA protesters calling up ChatGPT queries to set on fire at the barracades

FINISH RECONSTRUCTION

*peter sagal voice* ok panel now that Roberson is out of the running who will be the *next* member of Brandon Johnson's administration to probably lead the CTA

article: LLMs are unthinking inhuman text extruders and we should not anthropomorphize them me: strong start, hard agree article: so i recommend cussing when you ask them to plan your trips me: what

man said put it on my tombstone

"The governor of California is the host of a conservative podcast" is a simple literally true sentence

You can really see this cop deliberately take aim at the reporter while she talks to the camera They are often intentional like this, but it's not always caught on video so clearly

"it's not peaceful for people to be on the highway" -LA Mayor Karen Bass right now. This is what I mean when I say the word "peaceful" as applied to protesters is about prescribing placidity, not nonviolence. Calmly marching on a freeway is pretty inarguably nonviolent. " Peaceful" you can debate

full blown police riot in progress. never let them reframe this.

There is also a fundamental inconsistency in the premise of such tools. It means you think 1) Students shouldn't trust AI to do their work 2) Teachers should trust AI to assess the authenticity of student work

This is a great episode

Reduce stranger murders by a third with this one weird trick

talking with @miriamcb.bsky.social reminded me of a project I did a while back mapping Chicago's school attendance boundaries. My goal was to identify what I call "zone proximity conflicts", places where the zoned school is not the closest school. There are a lot! publicdatatools.com/zpc_map.html

talking with @miriamcb.bsky.social reminded me of a project I did a while back mapping Chicago's school attendance boundaries. My goal was to identify what I call "zone proximity conflicts", places where the zoned school is not the closest school. There are a lot! publicdatatools.com/zpc_map.html

we were recently asked in an interview how we handle the stress of ~waves hands~ and i think about how @anthonymoser.com makes music, tools and jokes and how @aoife.bsky.social plays sports, rides buses with me and takes photos and i just love that we have each other and art in our lives ♥️

I'm always talking about how much AI sucks In the words of the great @levarburton.bsky.social, "But you don't have to take my word for it"

one reason for this is how few US newsrooms are willing to interview and hire immigrant and international journalists due to legal costs for work visas, and how often non-white immigrant perspectives are considered 'bias' instead of expertise

You know what else generates probabilistic answers that are sometimes true? when somebody says, "we just need to improve the accuracy" or "we can fix the hallucinations," you should think about the magic eightball An LLM is fancier, but it has the same fundamental relationship to truth

"So what do you call this act?" "The Ethicist!"

I think LLMs / associational models could be useful in a context that centers human judgment and uses association to surface source material the reason we don't see those approaches with "AI" companies is that "AI" isn't about the technology. It's a political project to centralize power