mia

🏳️‍⚧️ she/her — without each other, we are no one — patchstep . com on bsky

[mia @patchstep]

📌🧵 2026 dailies

[mia @patchstep]

why do i have more followers than the ceo of opencode

[maria @maria_rcks]

Why is the CEO at @opencode the least followed person in the company?

[mia @patchstep]

mmmpf...
pickle for the west

[mia @patchstep]

@app_settings @luciascarlet its so dump i watch on 480p constnatly on mobile

[mia @patchstep]

i cant get it to generate anything that doesn't violate the guardrails around nudity, sexuality, or erotic content.

[mia @patchstep]

oh

[Theo - t3.gg @theo]

What the fuck

[mia @patchstep]

i promise i have not asked and do not ask chatgpt for anything problematic

[mia @patchstep]

@batuhan ohhh okay that makes sense. hi anyway

[mia @patchstep]

@batuhan WAIT I DIDNT KNOW BEEPER WAS DUTCH

[mia @patchstep]

@batuhan bunq user spotted

[mia @patchstep]

@emanueledpt @interaction telegram because beeper

[mia @patchstep]

baseball caps are just brim-challenged hats

[mia @patchstep]

Jubilee Surrounded: 1 Claude Mythos vs 20 Claude Opus

[mia @patchstep]

@Porkbun @AnupamHaldkar can't wait! fingers cross .ss is also on the list, so i can centralize all of my domains on porkbun

[mia @patchstep]

https://t.co/qMz3tQQvhL hit 10k visits earlier today and i missed it. who is even using this?

[mia @patchstep]

this seemed fun but i figured there was more to get out of this so i gave it a try.

[mia @patchstep]

update: poke asked me to apply for a job at @interaction

[mia @patchstep]

they did well on personality i suppose

[Poke @interaction]

Say hi to the new Poke! 🌴

Now officially approved by Apple to text on Apple Messages.

As the first and only AI agent. Chat now: https://t.co/VIWYU64dUI

[mia @patchstep]

@youne_2 @Retr0_Rider @Gee_7556 you mean myrrh?

[mia @patchstep]

@WalmartCAGaming i am patch

[mia @patchstep]

God I fucking love data modeling
Why do I love this so much?

Am I just a control freak and have an obsession with categorizing and disambiguating?

[mia @patchstep]

So this evolved a little and I’ve modeled a pretty expansive taxonomy and data model to represent different forms of art and other culturally significant media in a four-tiered system.

mode/domain: how do we experience the work?
medium: how is the work materially expressed?
genre: what is the cultural significance and grouping of the work?
work: concrete identity.

works can be multi genre, multi medium, and multimodal, and every tier can relate to multiple of the tier above.

And each tier has intra as well as inter relationships (genres can relate to other genres, as well as to authors and works)

All tiers will have a relationship to Patch as an interest with different relevancy vectors like ratings, frecency, and more.

[mia @patchstep]

my brain is going crazy with ideas. now i'm working on a relational database of "interests" that she can consult to talk about relevant topics she's interested in. it'll be based on my own https://t.co/Bec7E85vFi and https://t.co/9fP9PvQAJS data with a frecency vector to inform relevancy, and "ratings" to let her know what she actually likes best.