Post

I’ve posted it before, but it feels evergreen The two hardest problems in Computer Science are 1. Human communication 2. Getting people in tech to believe that human communication is important

Jake Anbinder@jakeanbinder.bsky.social

I think we need an inverse thread where we talk about things that normal intelligent people universally agree upon that are for some reason hot takes within your profession

279 reposts1K likes

I’d add that a significant part of other “hardest problems” cited is the human communication “layer.” This has been true in my experience for naming, caching, and other technical problems. There’s the problem, then there’s communicating about it with humans, which can be harder than the problem.

Feels especially relevant now as most tech companies are dead-set on automating away whatever human communication still existed!

"The whole reason I got into this field was to avoid that. Where would YOU send me given the sole goal of minimizing my human communication?" - somethin' aome of them might genuinely say honestly...

To be honest, this goes for every specialisation in every field, technical or otherwise. It's just that the people in that field only see the issues with others because to them, they are communicating. It's just greek to the rest of us.

This is what I target as my role in tech - it is surprising to me every time how little time and effort teams will put into demos and documentation. "We dont need those contextual docs - other engineers will be able to determine the usage by its schema" Its not the code, it your feature goals!

when i was in uni the president of the cs student union argued that rape is good because it means the strongest genes get propagated. when i explained that this completely falls apart when you consider humans as social animals, he claimed that society wasn't a natural part of humanity

There’s always one guy… (Kidding. But every time I’ve ever posted this I get several instances of this exact comment. Or people insisting that the technical problems are harder. Or that the real problem is naming things… All of which I think is sort of an amusing self fulfilling prophecy of sorts)

Technically it should be 0. Off by 1 errors Because then you add in the implicit “counting starts with 0” But yeah we all try to automate ourselves out of a job, and we can’t do it