An interesting quote about using data to tailor their original content:

"And so, using the data to make funding decisions is a different game than using the data to tune the story, which we don’t do."

So, he's saying they won't allow data to influence a script once written, but they won't buy or fund that script if it doesn't fit the data. I'm pretty sure two things tend to the same outcome. In ML, you'd be accused of letting the model train on the validation set. Bad stats.