#Hi

1 messages · Page 1 of 1 (latest)

tulip shale
#

Next gen Interviewing idea

#

Obviously everyone is encouraged to use AI

grizzled flint
#

Agni can we do ranked gooning instead

tulip shale
#

But since every game is unique AI won't help you "solve" the game

grizzled flint
#

Not bad idea tho

tulip shale
#

Beyond just first principles

grizzled flint
plain ginkgo
#

I don’t like the competitive aspect personally

#

I think strategy games could be cool but they don’t need to be competitive

tulip shale
#

And you literally play ranked matches against them

#

This is just for OA

#

And then in interview you explain how you interated your algorithm

#

It's a super easy number for HR though

#

They can set thresholds

grizzled flint
#

So it’s the same game

plain ginkgo
#

Yeah leetcode isn’t uniquely cheatable

tulip shale
#

Like everyone with elo above 1400 advances

#

It's harder to cheat on this OA

#

Since you can't just say "Claude solve this"

plain ginkgo
tulip shale
plain ginkgo
tulip shale
#

Like there's literally no research on any of the games

plain ginkgo
#

Patterns will emerge

#

You can’t think of 1000000 unique games

grizzled flint
#

Idt this is viable due to the sheer time commitment either

#

OAs we’re min an hour

#

But like

tulip shale
#

For all your interviewing needs

grizzled flint
#

I don’t like 1 competition determining ur fate for the entire recruiting cycle

minor trail
#

I feel like this is only possible at high level roles like quant though like how many people are genuinely going to beat algorithmic gpt-5 solution

tulip shale
#

So you can compete again if you want

plain ginkgo
#

I like the attempt to think of a system that tests more relevant skills than leetcode monkeying but I don’t think this is it

grizzled flint
tulip shale
#

Idk, we're just trying to build the next gen of interviewing

#

And I think Leetcode is simply not it

#

You have to think big if u want to see change

plain ginkgo
tulip shale
#

The games could be simple too, so you don't spend too many hours grinding.

#

Like, you open up the OA, then get 3 hours to code something up

#

A simple game could be minesweeper but you're the person placing the mines

#

^ ideas like that

daring lily
#

np hard games only

tulip shale
#

Where it's super contained, easy to understand, but also non trivial

#

And evaluation is objective

#

Some people will clearly just be "ok I'll place all my mines in the corner"

#

But that's clearly not optimal