Denis Stetskov

~1,000 dead in France this week. The 400-euro fix is banned by facade rules. The same reflex that guards the facade stopped building reactors, batteries, companies. Europe is spending an inheritance it can't replace. techtrenches.dev/p/europe-run...

Watching everyone is cheap now. The only thing missing is a reason good enough to make it normal. Terrorism, children, safety. The reason is always real, and the freedom is always smaller after. The machine is built. A reason is always found. techtrenches.dev/p/the-cost-o...

A loud forecast costs the man on stage nothing. It costs a generation their craft. Every CEO who called engineers obsolete has quietly walked it back. Not one correction under the name that made the claim. techtrenches.dev/p/nobody-ans...

OpenAI guaranteed DeployCo investors a 17.5% annual return. On what? A business that hires humans to make AI work inside companies. The same OpenAI that spent two years saying AI replaces humans just underwrote a 17.5% yield on humans. techtrenches.dev/p/when-your-...

Fable 5 launched Tuesday. The US government pulled it Friday at 5:21 PM for every non-citizen on earth, including Anthropic's own employees. Europe spent 3 years writing AI regulations. The US just made them irrelevant with one letter. techtrenches.dev/p/europe-reg...

Anthropic's entire brand was a single binding safety pledge. The month it got expensive, they deleted it, then filed for ipo at $965b, and 3 days later said the industry should slow down, the rule was theirs when it sold. Once it cost them,it's the industry's problem techtrenches.dev/p/anthropic-...

Uber burned its AI coding budget in four months, then capped engineers at $1,500/mo per tool. It rebuilt the exact ceiling it spent a year tearing down. The bill never measured productivity. It measured who burned the most tokens to win a leaderboard. techtrenches.dev/p/nobody-won...

96% on Anthropic's stage Monday. 99% at Investor Day Wednesday. It went up before the slides changed. Broke down every primary source behind Spotify's AI coding narrative.What Honk actually does, what 15 years of infrastructure look like, and why none of it transfers techtrenches.dev/p/honk-is-no...

Volkswagen: $52B. 9 million cars. $320B revenue. Cursor: $50B. Wraps API calls around a model it didn't build. The economics say utility. The multiples say magic. techtrenches.dev/p/the-50-bil...

canvas got hacked through a free teacher account. 30 million students. patched it once, hackers came back a week later. put a ransom note on the login page. during finals. the entry point was a feature nobody thought to audit. it always is.

GM just fired 600 IT people and is hiring "AI-native developers" instead. They call it a skills swap. The people who know how GM's systems actually work leave. The people who know AI but have never seen the systems arrive. The knowledge just vanishes. Nobody notices until production breaks.

Shell shortage nearly cost Ukraine the war. FPV drones fixed that. Then Russia put them on fiber optic cable and they became unjammable. Now AI data centers are buying up the same fiber. Price went from $4 to $50/km. A Starlink terminal costs less than one spool. techtrenches.dev/p/silicon-va...

Palantir CEO is in Kyiv today meeting Zelensky. 100+ companies training 80+ AI models on real combat data. drone interception, target detection. While Silicon Valley argues about benchmarks, Ukraine is testing AI where failure means people die.

NVIDIA made $120B profit. OpenAI lost $14B. Same economy, two different bubbles. The infrastructure layer prints money. The application layer burns it. And every query costs compute, so scale doesn't fix the math. Who survives the correction? techtrenches.dev/p/the-ai-ind...

We trained AI on 518 million GitHub repos. Most of them are student projects, abandoned prototypes, and unreviewed Stack Overflow pastes. Then we were surprised that the output is mediocre. The model isn't broken. It learned exactly what we taught it. techtrenches.dev/p/ai-is-a-mi...

3 AM. Production down. Your name on the commit. Doesn't matter if you trusted the AI or validated every line until your eyes gave out. The incident report has your name either way. That's the senior engineer in 2026. techtrenches.dev/p/the-human-...

AI that actually matters: Michigan researchers trained a model to diagnose heart conditions from a basic EKG strip. Previously needed: expensive MRI, specialist, big hospital. Now needed: 10 seconds, basic equipment any ER has. Not another chatbot. Medicine that scales to where doctors aren't.

4x token throughput on NVIDIA's Nemotron 3 Nano. 1M context window. We've been testing inference costs across our clients' projects. Parameter count stopped mattering 6 months ago. Cost per token is all anyone asks about now. The "biggest model" era is over.