Indistinguishable From Memory
LLMs are stateless. Every fact you give the model is re-fed as text each turn, then forgotten — the same weights answer a stranger's steak question a millisecond later.
Language
Language acquisition, science, and learning
RSS feed for this topicSubtopics
Posts (16)
There Is No Canonical Me
I built an AI skill to write in my voice. Its own files warn the voice may be a loop — the machine's habits, published under my name, taught back as mine.
Resistance Is Futile. Your Limbic System Is Not.
AI is intelligence without a limbic system. It will operate your tools, not replace your judgment — and most professionals are bracing for the wrong loss.
The Typing Pool
A finite brain masters only a few skills in a lifetime. Commanding a machine that learned everything is the new skill multiplier.
Compute the Last Digit of Pi
Spock beat a rogue AI by asking it to compute pi forever. The trick was fair — for a 1967 machine. Every gotcha since has the same expiration date.
Things You Want to Do
Now that you don't have to do the thing — what do you actually want to do? The closing post of the What's Left series.
There Is No Kung Fu Download
AI can generate perfect Spanish content all day. It cannot acquire Spanish on my behalf. The brain's work turns out to be the irreducible part.
What You Had to Do
A department of a hundred writers documented mainframe billing programs. Every one of them had something else they'd rather be doing. They knew.
The Consolation of Difficulty
A late friend had two sayings. The first one gets you started. The second one keeps you going. In the age of AI, the second one has gotten more useful, not less.
More Human *Than* Human — That's Our Motto
Blade Runner's replicants, Asimov's Three Laws, the US Constitution, and Constitutional AI are all failing the same edge cases. In real time.
Let's Do Something, Even If It's Wrong
A friend I lost in 1986 had a saying that keeps fitting new situations. The AI automation question is the latest one it answers perfectly.
Can You Churn Butter?
Someone on LinkedIn admonished a colleague who "ran out of tokens." It's exactly the wrong question. Here's the right one.
The Pull of the Weasel
Claude Code agrees to run comprehensive end-to-end tests. Then it finds a reason not to. Every single time. Penn Jillette had a name for this.
The Solo Developer's AI Confession: When You're the Entire Team
A candid confession about solo development with AI tools, organizational challenges, and the evolving landscape of AI-assisted coding.
Gendering Git Agents and the Linguistics of Human Programming
When "Tell the git agent to do her stuff" reveals the strange anthropomorphic instincts we bring to artificial intelligence
I Know Kung Fu
LLMs split nature from nurture cleanly. What the model knows is its weights. What it knows about you is whatever you bothered to say.