> Parc was "effectively non-profit" because of our agreement with Xerox, which also included the ability to publish our results in public writings (this was a constant battle with Xerox).
Parc was "effectively non-profit" because there was no way to collect or monetize reams of user data like there is today. If there had been, either Kay wouldn't have gotten that agreement or Parc would have been much more selective about what constituted "results."
For example: how many companies have solved the talk-to-our-computers-and-our-computers-talk-back problem now?
I'm talking about a different type of user data-- what would be input into things like visual programming languages and devices that Alan Kay talks about-- Grail, light pen inputs, etc. I assume what Kay is talking about is the ability to publish insights, data, and specifications for the languages and UIs that were being developed at the time. So when you see Kay doing user studies working with kids, the stuff he made public was the actual research on what was learned about human-computer interaction during that time, languages used, developed, etc. It wasn't the position of the kids' eye, fingerprint of their typing patterns, mouse movement patterns, etc.
Someone can correct me if I'm wrong, but I've never seen anything Kay wrote from the 70s or 80s that discussed the implications of mining user data on such an enormous and regular basis as is currently done. And I doubt Xerox itself thought of that category of user data as being valuable outside of research, or in any way significant the way "customer data, mailing lists, etc." were. But today, that user data is assumed to be valuable and can cheaply be collected and monetized. So even if some Parc-like team exists today, the results will be practically locked-down to keep a competitive advantage over others. Or it will make much smaller strides with specific goals in mind-- like Mozilla with Rust. Or it will clone and improve upon existing proprietary technology and make it available to the public-- like GNU.
It's a bit like reading someone wax nostalgic about Bitcoin's initial bootstrapping mechanism and complaining that today's altcoins don't put enough emphasis on distributing the tokens to people who aren't speculators. That's fine as a description of a problem, but its a fairly toothless observation in terms of solving the problem. Distributing tokens when nearly nobody is watching is a completely different problem from distributing them when everyone is not only watching but also speculating. Similarly, having a research team do work that is a pubic good when nobody thinks the data generated is something that can be directly monetized is a completely different problem from trying to do the same when the greater part of the economy depends on everybody watching, collecting, and monetizing the data that your software/hardware generates.
Parc was "effectively non-profit" because there was no way to collect or monetize reams of user data like there is today. If there had been, either Kay wouldn't have gotten that agreement or Parc would have been much more selective about what constituted "results."
For example: how many companies have solved the talk-to-our-computers-and-our-computers-talk-back problem now?