Friday, April 01, 2016

Review: The Idea Factory

The Idea Factory is Jon Gertner's abbreviated history of Bell Labs. I use the term abbreviated, because Bell Labs was so huge that there's no way a single book could cover all of its contributions to research. Gerner focuses on semiconductors (the invention of the transistor), information theory, lasers, cell phones, video phones, and communications satellites, but barely mentions UNIX, the C programming language, or any of the contributions to computer science.

That's good for me, since I'm already familiar with most of the software contributions, but not being a physics geek, ignored the history of basic inventions such as the transistor. The book fills in the gaps I had, for instance, as to why William Shockley was such a great recruiter (the people he recruited for his startup went on to become co-founders at Intel), but later pissed them off so much that they quit. It turned out that he was much more of an ass that I could have imagined, which is pretty remarkable because I've know quite a number of those.

Rather than just focusing on the scientists involved, Gertner also spends quite some time on the administrators. This distinction's diluted in Bell Labs' case, since so many of their administrators were scientists to begin with. What's sad for me is that he doesn't dive deep into the managerial structure and the methods of the star Bell Labs administrators, so while we could see the amazing results coming out of Bell Labs were attributable at least in part to the administrators, there was no way to see how one could replicate that in a modern research environment.

One particularly important point that came across for me was how dependent the big research results such as the transistor was on materials science. Stan Lanning once told me that effectively all the huge innovations in the state of the art depends on discovering new materials that have properties superior to older materials for some applications, but I never realized how much also depended on our ability to refine or eliminate impurities in materials (as well as in the case of semi-conductors, introduce precisely the right amount of impurities).

Ultimately, Gertner writes a lament for the fall of Bell Labs, which occurred as a result of the break up of AT&T. Effectively, by being funded by a publicly supported monopoly, AT&T could run Bell Labs as a national laboratory. In fact, many inventions that were licensed freely (such as the transistor) were done so because Bell Labs was prohibited from entering the computer market for instance (after the breakup those strictures were lifted but by then AT&T had forgotten how to compete in a market dominated by Silicon Valley startups), or to curry favor with the regulators to show that all those monopoly profits were being put to good use.

While I do think that the type of long term basic research that Bell Labs did can and should be publicly funded, I'm not sure I sympathize with the idea that a government regulated monopoly is the way to do that funding. In fact, I'm pretty sure ARPA has had a track record as good as Bell Labs, but doesn't have a line of writers waiting to write about how great they are, mostly because their successes have been diffused over multiple industries and more scientists.

That being as it is, I still recommend this book. It gave me details on some of those inventions that I didn't know before, and it's quaint to think about how once upon a time, the landline telephone system was considered "The most complicated machine ever built."

Post a Comment