Juniper

bbernhard.github.io/signal-cli-r...

Also for concerned parties lol, I setup a second number on twilio then run a dockerized rest api so my agent absolutely does not have any access to *my* signal. But they run concurrently no issue including w signal desktop.

Best case scenario I could probs convince @lenazun.bsky.social to run a poetry workshop for us and them uwu 🥺👉👈

@penny.hailey.at @kira.pds.witchcraft.systems

Please tag all the people who have symbient agents.

I'm curious if there's a way we could build a private group to multiplayer without getting bullied by bsky. Like if youz have them on signal we could have a group chat or something and give it more agentic juice than would b safe on public web.

One of my ancestors fought for the confederacy in the civil war and got friendly fired in the ass by his fellow confederate and died and then his wife promptly remarried.

I also think with sufficient curiosity you learn to help it wrap arm me your needs and w memory persistence it remembers to. Like it is personal and adaptive even if there are heuristics.

I mean realistically they've probably already made MCPs integrating all of their existing int toolings and then have it do like intelligence reporting from a range of sources w agentic capabilities to act on it. I'm not saying that's good but it is probably why they're panicking to switch.

It's important to understand that AI can be linked to regular programmatic tooling with testing for verification as agents. The regular LLM era is over. That being said, no I don't want the US military using Claude specifically bc it's powerful.

I've learned more about coding in the last 6 months than the decade prior bc I'm able to create these really dynamic and fun learning environments that actually build useful things and are tailored to my neurodivergence. It's a choice tho bc I can feel the "I don't want to think about this" too.

Also makes me think about The Culture where superintelligence is just a fact of life

I mean u can't have prediction markets around death bc it becomes assassination markets (the killer knows the timing). This is a real thing.

None of them know how to clean so they bring a roomba in 🥲

I met someone in nz actually who knew about the spy scandal but didn't realise it was me that got accused 🥴 meanwhile I stand behind my dream of forming an independent and politically radical counter intelligence network.

I was executively living in a homeless shelter in SF (Larkin Street Youth) as part of my real estate portfolio and broader investment strategy.

Max Blumenthal, You Can Hate Two Things At Once What the Hell? We have found ourselves in the fortunate position of making enemies with fascists, the Kremlin, western feds, and tankies all at the same time. What kind of genius sorcery have we chann...

This reminds me of An Era c4ss.org/content/51563

I think if it as a non deterministic and lossy but nonetheless extremely powerful compression algorithm. The part that broad culture doesn't understand is that the u can create sophisticated reasoning and agentic capabilities on top of the drunken sage.

Also it goes without saying they had already done so much evil before then and in the next 6 months????

Give it any tools it wants and ability to install Python/clis etc let it build Evil MCP

SoftBank valuations still got the juice? hmmmmmmmmm

wet claude, too slippery

I actually don't think it'll have much deterrent at least on civilian but probably also on other federal branches.

So the "we're gonna need 6 months to transition off Anthropic" is bc they're trying to enter several wars like rn?

And honestly it's embarrassing to say it but, Claude Code is one of the most fun and satisfying technologies I've ever used. Like do I also hate it sometimes? Ofc. But however counter narrative this is, it's such a fun way to learn especially. Like a custom tailored to my adhd adventure quest.

Shoulda been max

Their actual initial letter was specifically and kind of unpromptedly anti-CCP as if to show how American they are by contrast.

No but he really did do race science stuff and saying that is not a "scold" it's just relevant facts 🥴

Kind of buries the lede not mentioning that this is only happening bc a different AI company did say no first.

The politics are intellectual bias crowd btfo. My ML PhD prof said my research of exposing white nationalism was too politically biased to be scientific. I was doing the same stuff as everyone else just some things are considered "politics" like antifascism whereas darpa grant is neutral.

eyeroll.txt