Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I like his articles, but the artificial constructs sometimes drive me up the wall. After reading through a fairly rudimentary strawman about outcomes defining the difference between obstinacy and persistence, we reach the last paragraph that trades a poorly defined word (persistence) for five poorly defined words (imagination, focus, energy, judgement, resilience)

persistence is also defined by flexibility in thinking, appetite for risk/comfort with uncertainty, low ego. equally useless

(I still love you PG, despite my dyspepsia)



I too like these Just So Stories of the Startup Underground, but damn, they come off like Malcolm Gladwell forming a taxonomy of the kids that got it vs the kids that don't.

These are all properties of people, that ebb and flow and change from each problem they are working on. It is more productive to talk about contextualized behaviors over the properties of people.


I've found that properties and other traits in people can be malleable and change over time.

I have tons of personal experiences where a new developer seems very obstinate because they've never had anybody really challenge the way they do things. They get onto a project and suddenly get put in their place by a more senior developer. It might have to happen once, or several times before they start to change how they approach things and become more humble over time.

But I agree, properties of people can change, behaviors you can change for a short time, but you inevitably will regress back to how you normally behave. As such, behaviors tend to be easier to observe and predict.


Good to make a conceptual distinction though.


Of course concepts are important and esp where they deviate from each other.

As a concept, it is important to understand human behavior and what motivates it, but I try and not brand people with a permanent attribute.


Someone clearly failed the marshmallow test as a kid (/s)

Yes, exactly. You put it better than I did. Shit's messy, non-linear, non-monotonic. No need to put a bow on it.


The marshmellow test was debunked. Turns out like many if these bad experiments when you factor socioeconomic status the test cannot be reproduced. Turns out poor hungry kids just tend to eat free food when its available.


HN downvoting scientific studies... I'm worried that this site has reached its eternal September. A huge loss since reddit went public and killed all real discussion, this was one of the last places with intelligent discourse.

https://anderson-review.ucla.edu/new-study-disavows-marshmal...




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: