Notes
Essays and notes on AI, design, and what I’m reading, watching, and building. Multilingual, with a few reading themes — read it your way.
Interactive guide · ~15 min · EN
System design, explained — one component at a time
A clickable walkthrough of APIs, databases, scaling, load balancing, auth and security — decoded from Hayk Simonyan’s full course. Tabs, diagrams, and a method explorer you can poke.
Open the guide →18 August 2026 · 10 min · EN · RU · ES · CA
The treaty covers the moon
American satellites were targeted and destroyed in combat for the first time this spring. The retiring head of the Space Force told the WSJ what that world looks like — and the one place the 1967 treaty guards absolutely is the one place he says a weapon would be useless.
Read →27 July 2026 · 11 min · EN · RU · ES · CA
Nobody was watching the physicists
A conversation with Eric Weinstein, whose argument is that America built the most dangerous capability on earth, stopped guarding it, and then let one group decide who's allowed to think. A note on gatekeeping, on Epstein as a science story, on a sky that closed over El Paso — and on which parts of it I could actually check.
Read →22 July 2026 · 8 min · EN · RU · ES · CA
What unwinding a dependency costs
Japan wrote its dependence on America into its constitution. Now that the guarantor looks unreliable, the bill is public: defence spending doubled, missiles 110 km from Taiwan, combat troops abroad for the first time since 1945 — and $550 billion still paid to Washington. Europe started asking the same question about AI in June.
Read →10 July 2026 · 7 min · EN · RU · ES · CA
The doomers wrote a happy ending
The team behind AI 2027 — the scenario where superintelligence arrives around 2030 and the rules never catch up — published a sequel on July 9. This time it's not a forecast, it's a plan: a verified US–China slowdown that delays superintelligence to 2040. For once, someone proposes closing the gap by design instead of measuring it after the fact.
Read →9 July 2026 · 8 min · EN · RU · ES · CA
The restaurant that loses money on every plate
A finance YouTuber lays out the bear case for the AI bubble in three moves: nobody really trusts the technology, the unit economics are broken in a way software has never been, and China distills the American frontier for cents and gives it away. Here's what's structurally solid, where it leans on people who are selling something, and the handful of signals actually worth watching.
Read →6 July 2026 · 9 min · EN · RU · ES · CA
A country you can't vote in
A More Perfect Union investigation traces a once-fringe tech-elite dream — replace governments with for-profit companies — from a 1997 doomsday book to a charter city in Honduras to a US-backed 'AI-native' zone in the Philippines. A note on what's documented, where the film overreaches, and why 'exit over voice' is the idea to actually sit with.
Read →30 June 2026 · 9 min · EN · RU · ES · CA
Made to fight sideways
A psyops expert says the left-versus-right brawl is the operation, not the news — keep people fighting each other and they never look up. A note on Chase Hughes: how attention gets engineered, why anger became the business model, and where the conversation turns toward consciousness and a line every ancient text repeats.
Read →22 June 2026 · 7 min · EN · RU · ES · CA
Nobody agrees what it does
In June the US pulled Anthropic's most powerful model off the market overnight. Then 350 security experts revolted. The fight isn't really about the model — it's that no one can agree what it can do.
Read →21 June 2026 · 13 min · EN · RU · ES · CA
Two fights we keep calling one
“Decentralized AI” hides two different arguments. Pull them apart and the whole fight — geopolitics, law, security — finally makes sense.
Read →21 June 2026 · 9 min · EN · RU · ES · CA
What “enterprise-ready” means — and when you don't need it
Michael Grinich built WorkOS to sell big companies the one thing that stalls great products: the boring security plumbing. A note on his conversation — what enterprise-readiness actually is, when it's a trap, and the new chasm opening under AI agents.
Read →7 June 2026 · 11 min · EN · RU · ES · CA
Mo Gawdat's three-year warning
The ex-Google officer is optimistic about the future — and not about the next year. A note on his Diary of a CEO conversation: what's hype, what's real, and what he says to actually do.
Read →