I liked Bell Labs ever since 1986 or 1987 or so. I was considering developing a C++ compiler, but was concerned about:
1. would I need a license to develop a C++ compiler?
2. would I need a license to call it C++?
So I called up their head intellectual property lawyer. He laughed, and said feel free, and thanked me for asking. He said other people just went ahead and copied things hoping BL wouldn't notice.
Probably an unrelated question, but why the "or even" caveat for OpenAI? They claimed to be an open non-profit, cynicism seems warranted at a level even moreso than Microsoft and Google.
You are right but for OpenAI there could exist in principle a strategy of beating the cynicism out of them by holding them to their word. I might have wanted to encourage people to be optimistic about that.
Elon might have stuck to his lawsuit, or someone like PG could come out strongly against Altman, but you might need someone of Mervin Kelly's stature and an army of lawyers to do it. Maybe Fei-Fei Li after she has earned her startup cred (so that the suits and politicians have the chance to take a technical leader seriously, put her on the board in place of Larry Summers, e.g.)
What's the ideal end acenario if someone does successfully smack them back down to stand by their original promises? And what's the worst case scenario if no one stands up to them?
It might be better for one's sanity to start looking at what people and orgs do instead of what they say or what others imagine/hope they do. It's a business and and the end of the day has to return all the money Microsoft invested in it. Ethics is what toilet rolls are made of at large corps.
Today, maybe, but I doubt you could have said that of Bell Labs, at least by Occam's razor. (Maybe they weren't a business all by themselves, but their output was effectively deployed to the parent concern by most standards)
While the current behavior likely is a consequence of the approach then. Their corporate lawyers don't want those legal aspects of the Unix wars, again and created their moat.
It's always interesting to read how many of the silicon valley startups were founded by people who had IP they'd developed at existing companies where they were told "we're not going to sell this feel free to take it and start a company". I feel like large companies are more reflexive about holding onto IP these days.
I was born in ‘88 so I’m envious of people who have stories like these. I’ve yet to experience working for the kind of leader seemingly prevalent back then.
AT&T's liberal policies regarding software (and computer hardware) weren't out of the goodness of its cold blue corporate heart, but one episode out of over a century of antitrust enforcement against the original Bell System and its subsequent incarnations.
Most specifically, the 1956 consent decree arrived at by the US Department of Justice. In part:
The consent decree contained two main remedies. The Bell System was obligated
to license all its existing patents royalty-free, and it was barred from entering any
industry other than telecommunications. As a consequence, 7,820 patents, or 1.3 per-
cent of all unexpired US patents, in a wide range of fields became freely available in
1956."
My understanding is that the decree also specifically* prohibited AT&T from engaging in either computer hardware or software markets. Consequence is that when AT&T developed operating systems, compilers, and associated software for the UNIX System beginning in 1969, it literally had no option but to give the fruits away gratis, or hold them internally.
That restriction was lifted after the 1984 anti-trust settlement, though it took AT&T a few years to start acting on that. Once it did the result was the "Unix Wars" of the late 1980s / early 1990s, which amongst other effects put a strong damper on commercial viability of BSD-based Unix systems, whilst an alternative coded-from-scratch run-alike system originating out of Finland faced no such headwinds. That system was Linux.
(This history is ... referenced in numerous sources, though I'm finding it difficult to turn up specifics. The 1956 DoJ settlement doesn't seem to be online at all, and this isn't the first time I've searched for it. Wikipedia discusses various elements of the AT&T antitrust actions, dating to 1913, but is fairly poorly sourced, see: <https://en.wikipedia.org/wiki/Bell_System#Kingsbury_Commitme...>. I'm pretty sure that the Unix / Linux history is covered in various sources, possibly some of ESR's writings or elsewhere. I'm pretty confident of my facts here, but would prefer being able to source my statements more concretely.)
This is all true of AT&T but remember Bell Labs was a subsidiary of AT&T, and by all accounts from the ground, the Bell Labs management were able to promote a culture shielded from the politics of AT&T.
(Compare with GoogleX, similar intentions, maybe better starting conditions)
IP issues-- an interesting case study would be how the BL managers handled Shockley.
1. would I need a license to develop a C++ compiler?
2. would I need a license to call it C++?
So I called up their head intellectual property lawyer. He laughed, and said feel free, and thanked me for asking. He said other people just went ahead and copied things hoping BL wouldn't notice.
So thank you, Bell Labs! You guys were the best.