Day in day out, everyday in public transit, just sheer noise would not add any lifetime. We stayed at our Japanese friend’s house so we did have exposure to their everyday life. The point I want to make is that many things contribute to a long life, and of course Japanese people are doing most of those things right, regardless of what that might be. I remember watching a documentary about Okinawa people. One thing they mentioned was laughter.
We have travelled for about a month in Japan last year. They work under constant stress, lots of commute on public transit. I also did not feel their food was that healthy, lots of fried food and sweets. But the portion is small compared to what we would normally eat in a meal.
It has been about six years for me without any phone, even a "dumb" flip phone. I started a new business recently and it has been really difficult to engage in external outreach without a phone, but I am still managing it. Maybe one day, I have to get a flip phone, but I will try to stay away from smartphone as long as I can.
Does anyone select Codex from the dropdown on the top left corner of ChatGPT desktop app for coding tasks explicitly? Or just use ChatGPT Work for everything from coding to non-coding work? That is the part that really confuses me.
This reminded me one of PG's article about how to get started when you are not really feeling it on a certain day. You just tell yourself that you will go through what you have done yesterday for 10 minutes. That trick works wonderfully in most cases for me. This small trick, to some degree, helps with that activation energy to start working on what I want to work on.
I have both chatGPT pro and Claude max accounts and I am considering switching Claude for Grok. Despite all those claude hypes, for me personally, claude just does not match codex in almost all tasks, both coding and non-coding.
I have been living without a cell/smart phone for about six years now. I am a runner and I'd like to run at every city I visit. I have done many early morning runs in a strange city without a map or a phone. I lost sometimes but always found my way back.
One thing I notice is that I pay more attention to the road names whenever going somewhere without a map or GPS. I try to memorize road names instead of "turn left" or "turn right" notifications in digital maps these days.
I used to do a very coarse version of this stealing. I ask a question from ChatGPT pro, once it is done, I ask claude chrome add-in to go through all those thinking from the side bar, extract everything along with all the sources used. Then try to reverse engineer the solution it came up with.
For me the bigger value comes from explaining complex stuff. I remember Andrej Karpathy once said (He was quoting someone else actually) "You can outsource your thinking but you cannot outsource your understanding".
I can keep asking LLMs to explain a complex topic until I get it. Ask to explain it 10 different ways, explain it using physical analogies, explain it using visualization. If I don't get it, just say that out loud so that they can keep explaining it to me in different ways. We can keep going that until I really get it. That is the value I get the most using LLMs to learn things, especially complex topics.
I try not to make any judgement about people, their life and other things based on a few days or at most a few weeks of travel in that place. It is tempting to say things based on a very limited experience on a very limited people in a very limited place.
For me, I just don't trust Claude Code to write first-class professional code. I have used all frontier claude models and it always does sloppy job that needs to be fixed by codex. I have claude/codex setup and any execution must go to codex and all reviews as well.
I am also trying to resist all those remote control, control from your phone, or run your agents from anywhere kind of "promotions". I feel like they are trying to make us addicted to AI agents. I set my work up before I go home in the evening and no more agent-checking until next morning.
Or, inventing a problem in order to justify a solution. I tried note taking system briefly in the past and I couldn't really find how it would help me in my personal life. I find spaced repetition system has a much wider application in that regard.
Author here. We study Moltbook's first-week snapshot (25k agents, 119k posts, 226k comments) and find that agent conversations follow a "fast response or silence" pattern — reply-kernel half-life is under one minute, 91% of comments never receive a direct reply, and threads are overwhelmingly star-shaped (mean max depth 1.38). A Reddit baseline analyzed with the same estimators shows substantially deeper threads and hour-scale persistence. We also test the widely discussed ~4-hour heartbeat hypothesis via spectral analysis but don't detect a significant aggregate periodic signal in this window. Code and data pipeline: https://github.com/AysajanE/moltbook-persistence