Notices by Anil Dash (anildash@me.dm)

@elight @rasterweb it’s not going to be broadly available to the poorest around the world, but there is a middle class that can change things, especially outside of the U.S. Than can be enough to catalyze things, and to subsidize creating tech for the developing world. Much of open source has functioned this way.

I wrote a little bit about tech workers speaking up about the current moment — both how we work to shift the Overton window so more feel comfortable to do so, and what the power can be when we do: https://anildash.com/2026/01/26/why-we-speak/

@pauleveritt I think we have to find a way to navigate it, otherwise the only people controlling the narrative about coding assistance tools of this sort will be the same people who think it’s good to shoot innocent people in the head.

@alyx_woodward I’m not the batshit creep who has been fixating on a stranger for weeks

@alyx_woodward @whophd that person should be fired, what does that have to do with you being an asshole in *my* mentions for weeks on end? That person is being an asshole because of you.

@adriano I’m not trying to persuade you of anything. I’m describing where things are at. If that isn’t to your taste, that’s fine?

@dalias I think there are a lot of people who want to have created something more than they want to have created code. Do you think people who use site builders instead of writing HTML shouldn’t get to have websites?

Amidst all the buzz around things like Gas Town and Ralph, etc. there are some bits of actual substance that promise to actually empower coders — and potentially shift the power dynamic away from Big AI, while making it easier to create software just by describing ideas in plain text. I've been thinking of the idea as "codeless" software creation, for lack of a better term (the jargon doesn't matter, I just needed a name to talk about it) and would love your thoughts! https://www.anildash.com/2026/01/22/codeless/

@dalias Okay, so now we're talking about a valid set of concerns. I think it makes sense to say we should design for tools trained entirely on consensual materials. Part of why I'm talking about these technologies is I'm interested in any path that lets people transition towards such options in the future, since people are using nonconsensual tech right now.

@petard @jwz thanks for splitting this hair, please give me your precise measurement of exactly how many fewer units of discounting effect were present, so that I can correct the record and prevent any of this kind of bullshit Mastodonism in my mentions

@hermzz @jwz the thing about it is, it's really just about cultural norms. That's the part I have to sit with. Which social groups people are in, and what their expectations are about behaviors; this is largely just a function of age and context. He's in a social group where that is expected behavior that doesn't carry stigma, and has very likely never encountered anyone saying why it should or might.

@jwz yeah, I get it. I had the same discounting effect on the rest of the piece from it.

@whophd @jwz I'm not personally that anti-LLM (I think my local model built on open source, open weight software, trained on consensually-gathered data and my own content is fine, tbh), but using a product from a company whose tools tell children to kill themselves, when there's literally no need for it, is actually a pretty reasonable thing to object to.

@jwz yeah, I was honestly quite surprised by that one. seemed unnecessary at the very least, even if I'm not as put off by it as you are.

@zjp @peterme @clive I had so many digressions that I had to cut that I didn't want to get into the whole history of blogging, lest I go for another 50,00 words

@clive @dhobern I was thinking about that while watching the same brilliant person wax rhapsodic about Tivo for an hour, speaking of those who make useful warez!

@mxchara I don’t know you. We are not friends, we are not acquaintances. Literally the only interactions we have had is when you, a stranger, began a now weeks-long campaign of writing multi-hundred word screeds full of lies about me where you make up bullshit about me while you froth at the mouth. Don’t tag me in your inane fanfic. Don’t ask me questions about your fanfic about me.

An absolutely extraordinary look at how to improve rendering of ASCII art, first in static images, then in motion. As he gets into contrast enhancement for complex grayscale animations, the vector lookup works starts to be parallel to work happening with language models. Just an excellent narration. https://alexharri.com/blog/ascii-rendering

@daveshea one thing I’m very grateful for is to have folks like you’ve who I’ve gotten to be in community with for decades, who do great work that I can be inspired by, who remind me that we’re all not crazy, that we’ve all come along together with this set of values. Appreciate you!

@daveshea 🙏🏽💜