# All the work at Bell Labs needed to tie back to the system There’s a common conception that research at Bell Labs was completely unfettered. My hunch is that this was not the case and all the work at Bell Labs did need to tie back to AT&T’s core business in the same way that any corporate lab does, but the nature of [[“The System” of AT&T]] and it’s particular location in time and place enabled many researchers to forget about the box that they were in. AT&T’s system was so vast and multifaceted that tons of research *did* tie back to it - chemistry, electronics, information theory and other mathematics, materials. [[Bell labs brought many disciplines together under the same roof and broke down barriers between them]] not because abstractly that’s how you get good ideas but because that is what they needed to do to improve the system. It also helps that this wide range of relevant work happened to coincide with the most exciting research fields at the time. There’s *heavy* mixing between correlation and causation here (silicon electronics and information theory were basically created at Bell Labs, etc.) However, it is important to note the convergence between hot research topics and things that were actually relevant to AT&T’s core business. Today, this type of convergence is arguably a big reason why [[AI research is an exception to the decline of corporate R+D]]. [[Bell Labs management was extremely light]], which presumably helped with the perception that the research was unfettered. Of course, management could only be so light at least in part because the range of relevant research was so large. Thanks to the size of the boundaries on useful research and the quality of the management, I suspect that they were also able to hire people who had no desire to leave the box. It’s easy to feel unfettered when given free rein, you would do the things that a manager would want you to do anyway. This is of course the ideal everywhere but the circumstances at Bell Labs made it easier to achieve.[^1] [[Claude Shannon]]’s experience illustrates what happened when researchers strayed outside of work that wasn’t relevant. [[It’s hard to know exactly where a line is until you cross it]]. These anecdotes look like something that would happen at any corporate R&D lab. While there’s no concrete evidence, I suspect that incidents like this happened not-infrequently, but they happened with less prestigious people so they weren’t remembered or recorded.   ### Related * [[Expensive research needs to address an existential threat eventually at an organizational level to maintain support]] * [[Innovation orgs need to be aligned with their money factory]] * [[Regulation drove many of Bell Labs’ actions]] * [[Bell Labs did not give researchers complete freedom]] [^1]: Sidenote - I suspect this is the case with [[JCR Licklider]]’s program at ARPA. [[Alan Kay]] describes him as giving researchers free rein, but zooming out, this was not universally the case. Instead, it seems like Licklider was clear enough on his vision and what needed to be done that he could find researchers who would happily stay in the box he wanted them to be in. <!-- #evergreen --> [Web URL for this note](http://notes.benjaminreinhardt.com/All+the+work+at+Bell+Labs+needed+to+tie+back+to+the+system) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/All+the+work+at+Bell+Labs+needed+to+tie+back+to+the+system)