bickov

The per-app breakage is the real problem with text extraction, not accuracy, like etchalon pointed out. Screenshots dodge that since they look the same no matter what is rendering. Cropping to the window and marking it up before it goes anywhere also skips the full screen recording and the orange icon. https://slimsnap.ai already works that way.

[flagged]

The tiredness isn't from being in the loop. It's from what the loop hands you. Reading a wall of an agent's prose to check if it understood a screen is exhausting. The same check as a marked element with coordinates takes a second. It's not a human-attention problem, it's format problem. Models emit prose because it's cheap to generate, not cheap to review.

Says a lot that "how do I copy text out of it" is a solved-by-secret-command problem. Great agent, the terminal UX is held together with /commands

Banning the PRs is the easy part. The funny bit is that "understands their own code" is now a filter worth writing down.

Nobody misses boilerplate. What's gone is that understanding used to be mandatory, you couldn't ship what you didn't understand. Now you can, so it became optional, and optional things atrophy.

yes, but reader is changing. A lot of the time now it's an AI bot scanning the blog post and deciding whether to suggest it to human. So I write to be scanned , not just to rank. Same value, new format.

Moonwalked http://coub.com/view/d9ncc

Curated copy inspiration with the best examples of influential writing from across the web to help you take your text to the next level.