--- title: Why are flights productive date: 2026-11-05T18:02:41+05:30 categories: - how-i-do-things description: I used to think flights were productive because I had no WiFi; a connected flight changed my mind. Being stuck in my seat for seven hours may matter more. tags: [productivity] --- This Thursday, I flew Indigo from Singapore to Chennai. I had no WiFi and got a lot of (slightly difficult) reading done, such as: - [Machines of Loving Grace](https://tareqistyping.com/prototypes/reading-long-form/) - [Against the Personal Theory of Agents](https://www.gettheleverage.com/p/against-the-personal-agents-theory) - Paul Graham's [The Top Idea in Your Mind](https://paulgraham.com/top.html) - [How I Built the Most AI-Pilled Company in the World](https://x.com/jsnnsa/status/2115623382721126826) - Catching up on [Simon Willison's blog](https://simonwillison.net/) - Actions from a long pending book cluster summary (The Beginning of Infinity, Conjectures and Refutations, The Logic of Scientific Discovery, The Fabric of Reality, and Enlightenment Now). - Several tweets, blog posts, and ChatGPT / Claude chats. - I also drafted prompts to run later. In a disconnected flight, I'm less distracted because: - I don't need to check on whether an agent completed. - I can't open Hacker News if an essay is hard to read. - I can't click on a link in the essay. That forces me to absorb _harder_ stuff. Last Thursday, I flew Emirates from Brussels to Dubai. I had **good (free) WiFi** on the flight. Strangely, I _still_ got a lot of things done: - [Created an Asterix browser](https://www.s-anand.net/blog/reading-asterix-from-brussels-to-dubai-via-singapore/) - [Created a comic story using real faces](https://www.s-anand.net/blog/using-real-faces-in-comic-stories/) - Ran an agent-swarm experiment - Improved video embedding on my blog - Published a few blog posts It's a different _kind_ of productivity - doing rather than understanding - but still felt more productive. I think that's because: - I didn't have any meetings - unlike typical weekdays. - I can't snack or watch TV - unlike typical holidays. I tried this by sitting in libraries, at office on holidays, at cafes and client offices. It's not the same. There's something about the forced "You can't get up for 7 hours" that makes me a bit more productive. Your mileage may vary. But for me, flights are productive. I thought it was because there was no WiFi but it may be because I can't get up. ![](https://files.s-anand.net/images/2026-11-05-why-are-flights-productive.avif) --- A day later, I watched myself as I got distracted. Here's what I noticed. - When tasks are running in parallel, I get distracted. "What still running? Did it complete?" - Grouping all live tasks into a bucket - e.g. a tab group - helps. I'm peaceful there's one place I can go check. - Hiding the live task list helps - e.g. the browser vertical tab list and VS Code terminals. It's more work to check what's running. - When something updates on the screen, I get distracted. "What's that?" - Turning off things that update the screen helps, e.g. live sessions, notifications, even the running icons in browser tab / VS Code terminal titles - Two screens have more distraction opportunities than one. When do I get distracted? - When I get ideas (meetings, reading news), I get distracted. That's a good thing - but I need to complete them and that requires delegation, prioritization, etc. - After finishing something (submitting a prompt, committing code), I get distracted. That's a good thing - there's no point waiting for an agent to complete, usually. - While doing other things (exercise, TV, brushing teeth, etc.) I get distracted. Up to a point, that's OK. - When work is hard (complex, boring), I get distracted. That's a bad thing - and precisely when I need to focus. This is not structurally different from my [fairly detailed experiments in 2022-23](https://www.s-anand.net/blog/zeigarnik-effect-vs-my-procrastination/). I haven't consciously applied those learnings to AI agents, I guess. Worth practicing.