As someone who just left academia for industry a big factor in my decision is that I'll actually get to do more of the fun parts of research in industry. Ie de

As someone who just left academia for industry a big factor in my decision is that I'll actually get to do more of the fun parts of research in industry. Ie develop a system, test it in real world situations and then see it pushed into actual use within the span of a few months and be part of the group that benefits materially from it. By comparison the cycle in academia would be something like come up with an idea, submit grants wait months to hear back while working on other projects, do the work, write it up and submit it somewhere then wait months to hear back again and hope someone reads it and finds it useful. I think this also points at part of the solution: less of a wall between academia and industry and allow people to move more freely between the two. Machine Learning and data analysis is leading the way here with great publications coming out of industry and professors holding positions at companies etc.