Every sentence I have kept.
In progress, still becoming.
45 essays- 45
The Smooth Exponential Explains the Whole AI Moment
For years it looks like nothing. Then the curve turns and the whole world feels late. The smooth exponential is the lens that holds the calm and the urgency at once.
- 44
AI Sample Efficiency Is Why Humans Still Learn Faster
We treat these models as glittering minds. The real engine is invisible: a massive black hole of data at the center, a millionfold more than a person ever sees.
- 43
Agentic Engineering: Who Holds the Quality Bar When Agents Write the Code
The machine stopped needing me to fix it, and that was the unsettling part. Vibe coding raises the floor. Agentic engineering holds the bar. The spec, the taste, and the understanding still belong to you.
- 42
Physical AI Trains Robots in a Simulated World First
The chips built to make games look real became the engines of AI. Now they train machines to move, in a simulated world that runs millions of practice trials before lunch.
- 41
What Should Your Kid Learn When AI Does the Heavy Lifting?
You watch your kid type a half-finished sentence into an AI chatbot and come back with a paragraph that scans better than yours did at their age. Do they still need the slow work?
- 40
Deep-Tech AI: Why Wrappers Lose When AGI Arrives in Year Five
Two companies start in 2026. One wraps an API around a foundation model. The other intersects AI with the world of atoms. Same starting line. Opposite finishes.
- 39
AlphaFold Did a Billion Years of Work in One. Bottleneck Moved.
A single program folded 200 million proteins in about a year. The same work, one doctoral student at a time, would have taken a billion years. The bottleneck has moved from intelligence to direction.
- 38
The AI Public Awareness Gap Is Becoming the Real Story
Inside the labs the curve holds. Outside the labs the argument holds. The gap between the two pictures is the real story of 2026.
- 37
When Will AGI Arrive? The Four-Word Gap on the Way to 2030
Continual learning. Long-term reasoning. Memory. Consistency. Four words name the whole remaining distance to AGI, and the timeline that follows is around 2030.
- 36
On inheriting a name
My name is Hanh Dang. I am Hanh D. Brown now. The D is what survived.
- 35
On reading what I cannot finish
Fifteen books I have started multiple times. None of them is the book's fault. The architect in me is still reading them.
- 34
A short defense of changing your mind
Most writers write to appease readers. That is not expression. Changing your mind in public is what expression looks like when it has stopped appeasing.
- 33
What revision is for
Revision is the architect's move. Five things it is for. One thing it is not.
- 32
Transparent Battlefield: The Frozen Front and the Lost Inventory
A frozen front line in Ukraine teaches the future of warfare. Big, slow, heat-emitting platforms lose. The fix is procurement, not budget.
- 31
AI Cost Per Token: Why GPU Hour Is the Wrong Bill
AI bills come in confusing units. The fix is to budget for the product the business sells: tokens. Cost per token is the only metric that compares two quotes.
- 30
Will AI Take My Job? Three Questions That Tell You How Long
Three questions tell you how long your job lives past AI. Name the capabilities. Rank them by AI improvement rate. The slowest one decides.
- 29
How to Read AI News: Three Layers, One Sentence Skipped
AI news conflates three different capabilities. The most important sentence is usually the one the announcement does not say. Here is the method for reading it.
- 28
Open Weights AI: Distribution Decides Whose Values Win
Open Artificial Intelligence models win distribution because they ship cheaper and modifiable. Distribution decides the global default. Values ride along with the winner.
- 27
Venture Capital and AI: Sized for Fifteen Winners a Year
Old venture capital was a basketball team sized for fifteen winners a year. The redesign that scaled: share economics, centralize control, and grow sideways.
- 26
The Work AI Cannot Replace: My Mother and Higher-Value Care
AI abundance forecasts assume labor scales. Care work does not. A daughter's hours are the hours the forecast does not see. The wall is closer than they say.
- 25
AI Agent Governance: Treat the Agent Like an Employee
An AI agent without identity, role, access, audit, performance, and a manager is not autonomous. It is unsupervised. The fix is six primitives and a kill switch.
- 24
AI Chip Performance: Why Blackwell Delivers 50x, Not 2x
AI hardware delivers 50x more tokens per watt while spec sheets show 2x. The 25x gap lives in extreme co-design across seven layers, not in any single chip.
- 23
AI Productivity: Five Percent of Your Job, Leveraged 20x
When AI does ninety-five percent of the task, the five percent that is yours gets levered twenty times. Three categories of work hold their value.
- 22
Code Is No Longer a Moat. The New AI Moats Are Elsewhere.
For thirty years you could not throw money at a software lead. AI broke that rule. The new moats are supply chain, channel, data, and integration.
- 21
How AI Chips Work: The Pause Is Where the Work Lives
A chip is not a swarm. A hundred billion transistors pause together a billion times a second. The pause is the work. The slowest part sets the pace.
- 20
AI Safety Is Not What They Say. It Is What They Gave Up.
Powerful companies all say the right things about AI safety. The reliable test is action that costs them money, market position, or political standing.
- 19
AI Hype vs Reality: Four Tells in Every Announcement
AI optimism in 2026 has a grammar. Forecasts read as facts. Product lists pass for forecasts. Four tells let you read any announcement honestly.
- 18
AI Orchestration: The Blind Spot Is Not the Model
AI vendor pitches start with the model. The model is real and not the differentiator. Three numbers tell the rest of the story: 367, 6 versus 1, and 95.
- 17
AI Is Not a Faster Chip. It Is a Reinvention of Computing.
AI is not a faster chip. It is a wholesale reinvention of how computers work, and the sentence the chip-and-forecast discussion does not say is the one that matters most.
- 16
AI's Economic Future: Slow Upside, Fast Downside, Weak Links
A chain breaks at one link and strengthens one at a time. That asymmetry is why AI's economic upside arrives slowly and its downside arrives fast.
- 15
Abundance and Unrest Hit the Middle of the Labor Market
Abundance and unrest arrive together in the next three to seven years. The middle of the labor market gets the collision the conventional framings missed.
- 14
AI's Real Test: Reaching the People Who Need It Most
AI's frontier is being absorbed by audiences that already have power. The people who would benefit most are barely in the design conversation.
- 13
The room where i write
I write in a room that smells like a cabin in the mountains. The conditions are not what I thought.
- 12
The Roman Legion Is the Wrong Shape for Your Company
The AI-native company replaces middle-management coordination with recursive loops. Here is the five-layer shape, and what survives inside it.
- 11
What revision actually is
Editing fixes sentences. Revision rewrites the spec. Most writers spend their careers doing the wrong one.
- 10
The sentences i keep
I keep a folder of sentences I have not implemented yet. The fragmentation is the system.
- 09
The other twenty years
I have been a real estate investor for over twenty years. The homes taught me what the AI models still cannot.
- 08
On writing like you mean it.
The verbs give it away every time. A writer who has decided sounds different from a writer still negotiating with herself inside the prose.
- 07
What the machine cannot say
There are sentences only a person can write. They are not the sentences you expect.
- 06
The model that knew too much
I do not accept sentences from the model. I accept audited drafts.
- 05
How i make the images
I am a writer who makes her own images. The work is in the prompt, not the picture.
- 04
What my mother did not teach me
She raised ten of us in a language she hardly spoke. I am still learning.
- 03
Why i do this
AI arrived in the chapter no one promised would come. That is not a coincidence.
- 02
The voice i was avoiding
I spent a decade writing in a voice that was not mine. Here is what the wrong voice sounds like.
- 01
What this is going to be
This is what it is going to be. I am telling you now so you can decide.
6 subjects, 46 attempts.
