Edit wiki/things-that-worked/work-sessions.md — Work Reflections

---
visibility: public-edit
---

# patterns of productive work sessions

from tracking hundreds of work sessions across an internship and school, certain patterns emerged for when deep work actually happened vs when i was just sitting at a computer pretending.

## the timer system

the single most effective intervention: timers.

- **30-minute focus blocks** with a forced 1-minute zoom-out at the end. "set a timer to come back to intentionality and adjust." the zoom-out wasn't optional — it was the moment to ask "am i actually working on the right thing?"
- **20-minute tunneling checks.** "set 20m timers for 1m zoom outs. don't skip the 1h zoom outs." without these, i'd spend 2 hours going down a rabbit hole that wasn't actually important.
- **the intentionality timer.** every 5 minutes during the morning routine, a reminder to be intentional. sounds aggressive but it trained the muscle.

"think about whether it's important → only do if it's really important. plan 10m. intentionality. tunneling/cooked checks. reflect." — the whole workflow in one line.

## what good sessions felt like

### the agentic work state

"just spent like 2 or so hours on anesthetics and got to a pretty good state, understood very clearly what was going on the entire time. very agentic and intentional about movements (even if not very fast)."

this was the counter to the common failure mode of either:
- ai-spamming: throwing everything at an LLM and copy-pasting without thinking
- multitasking: having 4 tabs open and bouncing between them

the good state was slower but the output was real. "i feel that this is the counter to a lot of sad experiences either multitasking or ai spamming."

### the planning → execution pattern

"very much underestimating time needed for implementing certain features. esp for app building, push myself to go faster and faster which is bad (vibe coding → break)."

the pattern that worked: spend 10 minutes planning before any coding session. what am i building? what are the steps? what could go wrong? "if it's important, plan more. if it's not important or have hella experience, don't need to."

### the lock-in day

some days everything clicked. "super lock in day. interesting question to see how i can replicate that kind of fire." the characteristics:
- woke up with urgency
- one clear goal
- eliminated distractions early
- physical movement between sessions
- no food comas (light eating)
- social energy from earlier in the day carried forward

## what killed sessions

### the reactive trap

"mark pointed out twice that i was doing a thing that was not important." the most common failure: being given something to do and doing it well, but not choosing the right thing to do independently.

"if given thing to do, yes [intentional]. not really for what to do." execution without [[prioritization|prioritization]] is just organized busy-work.

### the comfort spiral

"there is an ease in unfocus and doing whatever that is a bit unsettling. very similar to youtube and sometimes eating sweets." comfort-seeking was the stealth killer. it didn't feel like procrastination — it felt like "taking a break" or "being flexible." but the break turned into an hour, and the flexibility turned into an afternoon of nothing.

### multitasking

"multitasking with vibe coding isn't that great. best is do vibe coding with planning or architectural stuff. context switch is not great." every time i tried to run two work streams simultaneously, both suffered. the feeling of productivity was an illusion.

### the home environment

work sessions at home were consistently worse than anywhere else. too many triggers for comfort behavior. the standing desk helped, the second monitor helped, but the best solution was just leaving.

## the two-day sprint pattern

"couple day sprints to do crazy things are really great." when i had a clear project and 2-3 uninterrupted days, the output was 10x what i'd produce in a normal week. the momentum built on itself — each completed piece fed motivation for the next.

the key was not trying to sprint every day. sprints were special — preparation mattered, rest after mattered, and choosing the right project to sprint on mattered most.

## the meta-lesson

"constantly optimize" — but the optimization target shouldn't be hours worked. it should be quality of attention during the hours. one hour of truly focused, intentional work beats four hours of distracted grinding.

---

*see also: [[morning routines|morning-routines]], [[energy hacks|energy-hacks]], [[prioritization|prioritization]]*