TL;DR Bring back Corporate Research divisions tax free and everybody wins. I've been a "research rat" all my life. I did research at Unimation, the first robot

TL;DR Bring back Corporate Research divisions tax free and everybody wins. I've been a "research rat" all my life. I did research at Unimation, the first robot company. I worked at IBM Research. I've worked at Universities. At IBM Research under Ralph Gomory, the company valued research. It measured itself against other companies like Bell Labs. I had the freedom to work on individual projects (e.g. rewritable paper and a Design-to-Build system that generated robot assembly plans from 3D CAD drawings). I worked on long term group projects, one of which was the Axiom computer algebra system. IBM valued "wild ducks", the people who thought of new things regardless of their place in the organization. For instance, the "nipple mouse" that used to be in the center of keyboards was invented by Ted Selker who worked in the math department. The key turning point seemed to be when Reagan shut down the Federal Labs. If you worked with someone outside of IBM Research your work was much more valued than inside efforts. The Federal Labs were a great partner in research. Once they were gone a lot of projects lost outside collaboration. The effect was widespread and deep. Axiom, for instance, was a multi-year pure research project in computer algebra. It was jointly funded by the company and the Federal labs. When the labs closed Axiom was sold, becoming a commercial competitor to Mathematica and Maple. (It is now open source and I'm doing unfunded research in program proofs for computational mathematics). Companies don't seem to invest in "real research". Apple, Google, Microsoft, and Facebook are sitting on billions of dollars of cash. You would think that they would create a research lab equivalent to Bell Labs (where the Neural Network work was done in the 90s), or IBM Research (where the atomic microscope work was done). Gomory used to say you could employ 3 researchers per 1 million dollars. A billion dollars a year is 3000 researchers. Gomory pointed out that only 1 in 10 research projects ever produced anything that could be sold. That would be 100 new products a year, assuming 3 people per project. Gomory pointed out that "real research" took 10 years to become products. So start now. There is a push to bring the "hidden dollars" kept overseas by Apple back into the U.S. at a lower tax rate. Instead, reduce the tax rate if the money is put into research. Let the big companies have tax breaks to run "corporate research", like Bell Labs and IBM Research. Then these company research efforts can work with universities to fund research graduate students and give them a career path worth their PhD. Tell Apple it can pump 1 Billion dollars a year tax free into a corporate research division and see how quickly the flowers bloom.