Ian Vanagas

When his friend and mentor Raiders owner Al Davis called Parcells to check in, Parcells relayed his injury issues. Parcell’s: “Al, I am just not sure how we can win without so many of our best players. What should I do?” Davis replied: “Bill, nobody cares, just coach your team.”

ML models are formed from combining biases and data. Sometimes the biases are strong, other times they are weak. To make a model generalize better, you need to add more biases or add more unbiased data. There is no free lunch.

DOOM was extremely humane software. It let you play with your friends, even in the shareware version. And along with the game itself, id released tools which allowed anyone to easily modify the game’s levels and assets, and then share them for others to use. This created a thriving ecosystem that exists to this day.

I turn down  almost  all of these Zoom invitations. Too  many  of them embody presentations, and I prefer more interactive modes.

By thinking about a ML problem first as a set of inputs and desired outputs, you can reason whether the input is even sufficient to predict the output. 

All the mental energy that you use to elaborate your misery would be far better used trying to find the one, seemingly impossible way out of your current mess.

A person with this perspective is one who witnessed the more orderly society functioning, yet did not witness the building of the more orderly society , and thus does not understand the mechanics of the more orderly society. In other words, this person has a distinctive blind spot —he knows what could be reaped from a more orderly society, yet not... See more