One thing that’s missed here IMO is the giant tech companies, which over the past ~15+ years have in fact been plowing incredible amounts of money into nominally unprofitable research. ChatGPT and friends, for example, originally came out of the somewhat-legendary paper “Attention is All You Need,” published by a group at Google. To say nothing of all the other stuff Google has funded (robot farming, internet from high-altitude balloons, various self-driving car projects)…
When we laud Bell Labs as an example of industrial research, I think people often elide the truly unusual incentives that created it:
- Corporate tax rates were high (~50%) and top individual tax rates higher (~90%), so there was no good way for company executives to use their excess profits to enrich themselves or shareholders. This makes corporate research investment much more favorable - what else is the company going to do with the money?
- AT&T in particular was a highly regulated monopoly, so it had large guaranteed profits and (as above) no great way to spend them.
- As part of its monopoly deal, AT&T was also required to license its patents to anyone for free. This meant there was less reason for Bell Labs to be secretive, and its discoveries were much more able to get out into the world. Nobody was charging royalties on the transistor or laser, or fighting in court to keep Shannon's work on communications theory secret.
No company today operates under these incentives, so in that sense Bell Labs really was a kind of "third space". I think academia is the better seed ground for this kind of work going forward, as awkward as the fit is.
One thing that’s missed here IMO is the giant tech companies, which over the past ~15+ years have in fact been plowing incredible amounts of money into nominally unprofitable research. ChatGPT and friends, for example, originally came out of the somewhat-legendary paper “Attention is All You Need,” published by a group at Google. To say nothing of all the other stuff Google has funded (robot farming, internet from high-altitude balloons, various self-driving car projects)…
When we laud Bell Labs as an example of industrial research, I think people often elide the truly unusual incentives that created it:
- Corporate tax rates were high (~50%) and top individual tax rates higher (~90%), so there was no good way for company executives to use their excess profits to enrich themselves or shareholders. This makes corporate research investment much more favorable - what else is the company going to do with the money?
- AT&T in particular was a highly regulated monopoly, so it had large guaranteed profits and (as above) no great way to spend them.
- As part of its monopoly deal, AT&T was also required to license its patents to anyone for free. This meant there was less reason for Bell Labs to be secretive, and its discoveries were much more able to get out into the world. Nobody was charging royalties on the transistor or laser, or fighting in court to keep Shannon's work on communications theory secret.
No company today operates under these incentives, so in that sense Bell Labs really was a kind of "third space". I think academia is the better seed ground for this kind of work going forward, as awkward as the fit is.