On Scott Adams. A man finds, to his astonishment, that he exists. After the elation of childhood wears off, he asks, who am I, why am I here, how does this work? These are hard questions, so after a brief struggle, he selects a readymade answer and goes about the motions of life. Scott Adams was not such a man. He was a live player, ever curious, intent on figuring out this simulation that he found himself in. From first principles, Scott unraveled, understood, and ultimately controlled his own reality. He hacked himself with affirmations, others with persuasion, the world with simultaneous sips. He explained people as moist robots, two movies happening on one screen, his world as Gods’ debris. He carved a personal mission to “be useful,” and made us all better writers, public speakers, and persuaders. He preached the footwear theory of motivation, the Adams Law of slow-moving disasters, the skill stack, systems over goals, and of course, the Dilbert Principle. Besides cartooning, philosophizing, and teaching, Scott rose to the occasion and displayed, “the one virtue that cannot be faked” - courage. Scott had the courage to speak honestly as he saw it - about Trump, about his nation, and about his time, even though it cost him friends, audience, money, and his ticket to polite society. Scott had true courage, the kind that makes you unpopular, the kind that is always and everywhere in short supply, At the end, as any hacker of reality, Scott covered all of his bases - he left as a Buddhist, a Christian, and a player in the Simulation. Scott, we didn’t get enough time with you, but you were a mentor and a marvel. You were useful and you were courageous. You were incompressible and indivisible. One of a kind, and generous with your drawing, writing, and speaking. Unlike your squealing critics in the chattering class, you will be read generations from now. On this earth there are many long-lived hells but no lasting heaven. Each heaven must be created and nurtured, ex-nihilo, from mind and from mud. Scott, you created a small heaven for us all, and to a larger heaven you go. A man finds, to his astonishment, that he no longer exists. He asks why, what it was for, and how will the new reality work? When the rest of us get there, we’ll find Scott, ever useful, ready to explain, having figured it all out. Notes: • First line paraphrasing Schopenhauer. • Courage quote via Taleb.

interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works already-clean.goodfire.com/

I have finally found a job where my “jack of all trades, master of none” schtick is actually pareto optimal and allows me to genuinely stand out. AND I get to use Claude to automate all the bullshit I am too lazy to do myself. God damn this feels good.

🚨 Manchester United players whose contracts EXPIRE at the end of the season. #MUFC [Sky, @ue_shm]

Yeah whatever man, you’re having a fuckin laugh 😭🤣😭🤣😭🤣🤣🤣🤣