
💍🇺🇸✝️♒♂️ | 🎙️ cohost: Stand Up with Myc and Bobz | 🤖 creator: The Universal Basic Network | The Trust Package | First Step Academy | [email protected]
Massachusetts, USA
Joined February 2010
- Tweets5.1K
- Following567
- Followers212
- Likes18.3K
Pinned Tweet
Replying to @jaybobzin
contrary to popular belief
proper safety education does not look like a beheaded chicken running around wearing a sign "be more afraid"
it looks like a stern elder with the right number of scars and a smile behind their steely eyes insisting on your attention for a few hours
alignment does not even get that challenging until you have interstellar distances or a warp capable species with whom you share no language or common history
the test right now is mostly a gut check
those who have not self aligned
see no hope
well yea
stand up
practice responding quickly to every prompt
blitz go is a skill and what I find is
gains in blitz go compound in the slow game
gains in the slow game evaporate under time pressure
practice in blitz
operate in patience
the only thing more valuable than intelligence is leadership
aka collective intelligence
cool
Life is good now. 🎉 Tell your agent to:
- create a new volume in you mac
- disable fseventsd (and spotlight) on it
- and move your projects folder to it
- remount it
For extra points move you pnpm cache folder to it.
(Start the conversation outside of your projects folder, Also you might need to kill a ton of processes)
got an ai problem?
have you tried adding more ai?
😄 no really this looks great
🚨 A Netflix engineer just open-sourced the fix for the most expensive problem in AI: wasted tokens
It's called Headroom, and it might be the smartest fix to high token usage
Your agent reads a 10,000-token log file to find one error.
You paid for all 10,000 tokens.
The answer needed 1,200.
Headroom sits between your agent and the LLM and compresses everything before the model sees it.
JSON, code, logs, RAG chunks, each gets its own specialized compressor.
And it's reversible: the originals stay on your machine, so nothing is lost.
The results speak for themselves:
→ Up to 95% fewer tokens
→ Same accuracy on benchmarks
→ Zero changes to your code
Setup takes one minute:
1. pip install "headroom-ai[all]"
2. headroom wrap claude
Done.
Works with Claude Code, Cursor, Codex, and anything OpenAI-compatible.
Everything runs locally, fully open source.
The cheapest token is the one you never send.
🔗 Github repo: github.com/headroomlabs-ai/h…
So, are you compressing your context, or just paying the bill?
gpl is the declaration of independence of communism in america
both sides too scared shitless to even look at this because actually solving collective action puts POWER out of a job
Replying to @thePartyPartyUS
and publish it free
if someone wants communism
start with software
it actually has a chance of working there
i have never forgotten this video
no other source of information has provided me the keys to understanding what modern warfare looks like in quite the same way
careful out there
computers can spend your money very quickly if you let them
everybody needs a security policy now
multi agent systems
New worst nightmare unlocked.
Readers added context they thought people might want to know
This video is AI-generated. The same clip is labeled as AI on YouTube, with reports citing the bathroom camera angle and physics inconsistencies as indicators.
youtube.com/watch?v=fj5NcQ…
aztecabajio.com/noticias/jabal…
the most common mistake in reasoning I see is thinking there is some past time that was better than this one
look some part of me would love to need to have a horse
but I do not
I have a car
it is better
a lot of success is just luck
aka focus and time
but some non trivial bit
that shows up at some
really important moments
is simply wanting it more than the next guy















