,
3 tweets,
1 min read
Read on Twitter
Prediction Machines fundamental thesis is that "predictions" will become a measurable quantity, w rapidly expanding supply, and thus a cheaper & cheaper input benefiting then transforming existing processes. Historical paradigms = indoor lighting, then electricity, then compute.
The book argues the complements of "prediction," ie things that either help AI gather valuable training data to make better predictions (eg sensors on a self driving vehicle), or people who need easier to understand prediction distributions, will most benefit from the rise in AI.
Contrast this passage from Prediction Machines with the guidance John Malone's father provided 
