In his year-end review, Hugging Face developer advocate Alejandro AO says 2026 was his most distracted year: three or four projects at once, and every one of them shallow.

Alejandro AO, a developer advocate at Hugging Face who makes videos teaching beginners to use AI, published a year-end review this week, and the headline was how distracted he had been.
Four to ten agents running at once
He had four to ten agents going at any time, pushing three or four projects. He burned millions of tokens a day and still shelved some of them.
AI makes it feel as if two lines of prompting will get a project done in an hour. So every idea that pops into your head gets its own build.
But the 'small' projects swell and swallow far more time than expected, shoving the important work further back.
Time does not get bigger
His conclusion is blunt: AI makes things easier, not time longer. A day is still 24 hours, and the more you run in parallel, the less each thing gets.
The goal is not to do many things but to finish the most important one. The better AI gets, the harder you have to resist the 'just build it quickly' temptation, and the more ruthlessly you have to cut projects and rank them.
He wishes for a 2027 focused on one or two core projects. It is probably a reminder to everyone: use AI to subtract, not to add.
Why it matters
When the cost of execution falls to nearly zero, the scarce resource stops being whether you can build something and becomes where your attention goes. The satisfaction of opening another project is an easy excuse for deferring the one that matters.


Curated from high-quality sources, with concise summaries and key takeaways.