When things go wrong in your company, nobody cares. The press doesn’t care, your investors don’t care, your board doesn’t care, your employees don’t care, even your mama doesn’t care. Nobody cares.
In the end, every single person like Nick who votes to deploy their capital into something which doesn’t produce anything beyond capital, is a vote to sit on the sidelines while every single other person works to keep the lights on, or ideally reverse the cycle.
The number of them that actually learn all the important stuff in under a month is zero. The number of them that have a self-guided strategy to learn what is relevant is almost zero.
For instance, I have been to conferences that have “speed dating” sessions (without the date part, to be clear, and with vaccine and testing requirements) where you meet many people for say two minutes and then move on to the next meeting. This should become a more regular practice.
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.
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.