Spot on. But I think there is also going to be a massive impact on hiring and teams. This ladies and gents, is where we won't need to hire people. Software 2.0

Spot on. But I think there is also going to be a massive impact on hiring and teams. This ladies and gents, is where we won't need to hire people. Software 2.0 could use the same infra, rely on 100 different datasets and produce 100 neural nets instead of a 100 teams. Unfortunately they wouldn't have the domain knowledge to understand most of the why or the domain. More examples of 1.0 that will be things of the past and could transition to 2.0: - people, entire teams or even companies writing branching logic for IVR (dialers) - almost all manually curated nocode drag and drop decision trees can be replaced with a NN. - ranking & sorting algorithms - prediction - ux (what to show first) - product recommendations - a/b testing - workflows with a bunch of ifs and branches 2.0 when you think about it is truly nocode. That said we are having a tough time visualising large neural networks. Radar charts work for < 10 weights. Tables are boring. Its time for innovating in visualising 2.0 as well.