
I spent the night at the Corgi cafe so you don't have to #airpartner

Asserting American biodominance 🇺🇸 for the betterment of the world 🌍 I don't do AI. I *am* AI. code: github.com/E-xyza

I spent the night at the Corgi cafe so you don't have to #airpartner

11 (3-5-3) (E-B-A)

Isn’t zero times negative one still zero?
Canada added no private sector jobs from 2019 to 2024 — U.S. did the exact opposite nationalpost.com/news/canada…

turns out nonprofit fundraising is as easy as... authoring a ballot measure to give your 3-person company $4 BILLION DOLLARS
Something feels a bit iffy about California prop 38. Doing a popular vote proposed by a 3 person non-profit that gives said non profit $4.2B from the public! I hope Gary is great and knows how to spend $4.2B well because it’s done.


like in 1 hour we have implemented the entire model, and tried like 5 optimizations. It's already 3x faster than python.

Claude having fun rewriting a transformer kernel in zig

I get it it's a pentest lite but my AIs are supposed to be virgin, innocent spirits

just wrote code for my agents to talk to each other. This was the test the coding agent wrote. fucking scariest thing I've seen in a year and a half of ai-assisted coding.

matt trudel doing incredible work rn keeping #elixirlang community safe

it's over, boys. star trek dominance defeated in opus 💀
When you ask Claude to pick random numbers, you really get the impression that it was trained on The Hitchhiker's Guide to the Galaxy.

I've come to believe that it's a lack of self-reflection, not being conniving or manipulative, that makes someone a bad person. Everyone engages in occasional behavior that is harmful. A "good person" takes that feedback to heart, takes responsibility for their actions, and tries to do better next time. A "bad person" pushes the thought away and refuses to reflect on their behavior. Over time, doing that repeatedly - refusing to be morally accountable - is what makes someone a bad person. It doesn't mean they don't have the capacity for redemption, but it's impossible to collaborate with people who refuse to learn from iterated moral games.

(it was opus)

Today the #elixirlang typechecker found an inefficiency in a jsonschema, and the LLM indeed fixed it by tightening the jsonschema. 😳

Unknown mechanism for your stage IV cancer experimental drug? Pill issue
not adapted for life in waterworld? gill issue

My fun side project proof checker was useful for understanding the OpenAI proof withdrawal: github.com/ityonemo/openai-w…