Hacker News Reader

Comment by marginalia_nu | Hacker News Reader

I've a hypothesis that a lot of the actual work is done in the background.

Focused thinking about a problem just sort of spawns new threads (some of which get OOM-killed if you keep spawning more). The process seems to need not actively thinking about the problem for the jobs to actually resolve.

It's like thinking about the problem triggers the background calculations to start, it doesn't actually perform the work.

Then you get that famous "aha!" in the shower or when you're off on a walk or whatever.

View on HN

The background processing can still happen while sleeping too (which is useful to use).

I went through the Navy's Nuclear Power program a lifetime ago and that was often 16-18 hours of instruction and study (and exercise) per day to do well. When you recognize that you're at that point of mental exhaustion you can sneak in something like a 30 minute nap to power through the remainder of what you need to achieve that day...

My general approach was to physically write down and do some checks to verify that I understand all of the material (reinforced memorization) and get as much in per day like a sponge.. and then sleep on it. The retention was nuts (was top of my class) and it kind of just becomes habit at some point.

I think the odder thing is that if you just keep doing it for decades you notice the missing information too. Those loose associations are still there for sure but the brain can only pack in so much at a given time or some such.

This seems to match my experience. I'll pile on a hypothesis of my own:

When we encounter a new problem, our brain spans a decision tree of assumptions from the closest known solution to a similar problem. The area around this tree is searched depth-first, which often brings good results (e.g. thinking about whether you'll put some threshold value at 0.2 or 0.25), but it's best at refinement. The more time passes and the more different things you do, the more the search is moved into breadth-first (which makes intuitive sense - the deeper the tree, the more processing power would have to be put towards it). Often times we literally cannot conceive a good solution to a problem, because our brain is focusing on finding the right deep leaf of the wrong subtree!

There's a really great book called Consciousness and the Brain that goes into all the various research around how our brain works, and there is a huge amount that we think of as "thinking" that is totally unconscious to us. As I've aged, it seems like more and more of my thought processes are unknown to me. I just get agitated thinking about things, go do something for a while, then my unconscious "me" is suddenly like, "ok, here's the understanding".

https://www.amazon.com/Consciousness-Brain-Deciphering-Codes...

The most productive hours at work for me has either been things I'm doing for the third time OR 15 minutes after I've left for the day after 8 hours of going in circles.

Privately, some of my most productive development has been when my girlfriend and I had separate apartments. I'd spend days working on problems undisturbed at my place, and then I'd go visit her for a few days (bringing no computer).

I always came back to my apartment full of inspiration, with half a notebook full of ideas and plans, which I'd spend the next days implementing, while spawning new background jobs that would mature the next time I was AFK. Rinse and repeat.

Something about this set-up was just insanely good for my productivity. I was much more productive than periods I didn't have these forced breaks away from input.

You just reminded me that I have also experienced a two location split week, with one side consistently in isolation.

It was a wonderfully creative & productive situation.

Indirectly this also betrays the kind of job where you had the "luxury" of being in a position to be able to take such forced breaks in the first place.

This is already a good situation to be with regard to burnout.

I'd love to do this (and try to when I can), but more often than not it's a balancing act between "rest but fall behind" vs "don't rest but catch-up".

While there's some element of privilege involved, I wouldn't have been able to ever work like this if I wasn't willing to take risks that many would consider too much. But the pay-off from taking those types of risks also (sometimes) ends up opening doors for more unconventional career steps.

I'm grant funded for the next year, and after that I have some savings to live off that may last another year or so (maybe less if this inflation keeps up), but after that I haven't got a singular clue where my money is coming from.

But I also don't have kids, and don't want any either. If I had or wanted, none of this would be possible. I'd definitely choose the golden handcuffs in that counterfatual.

Same. This is why it’s so important to go to bed late at night when working on a bug instead of grinding it out. I’ve found that when I wake up, I almost immediately have more ideas on what to do instead of beating the same dead horse I was at 3am.

This is similar to Rich Hickey’s talk on “Hammock Driven Development”:

https://youtu.be/f84n5oFoZBc

Spend your conscious time loading your brain with data. Spend your unconscious time randomly making connections… until “Eureka!”

This is why I don't bill hourly. Hiring me means installing your problem in my head, and that thing runs in the background, even while I sleep.

This is the basis for my personal system of working with ADD

The brain is actually rather active even when we think it's not. The parts that are active are called the default mode network

read Cleese's "Creativity"

Reply on news.ycombinator.com