I was in Bell Labs for 15 years, starting with the Bell System, and ending at the trivestiture.
First, talking about the BoCs like Pacific Bell has nothing at all to do with this thread. It was a big company. Duh. It did no research. When I was there more people worked for AT&T in New Jersey than worked for state government. In any case, phone books were really handy before the Web, and operators were really handy before voice recognition, so I’m not sure what you are complaining about.
The Bell System was high regulated. Overpriced - perhaps, but some of that was to support phone service to people that AT&T would never make money serving. A far more accurate complaint was conservatism, which came out of leasing equipment, and thus not wanting to “churn the market” as it was called. I went to a seminar about how to make phone equipment less reliable to meet the competitive challenge. Equipment back then was fundamentally less reliable than it is today, to get the kind of network reliability we had was no small accomplishment.
I voluntarily left what would become Lucent with a job at Intel and a nice piece of change. I can tell you that a lot of the job losses were not because of income but because Bob Allen wanted to cut the payroll to make the IPO easier. He wasn’t going to be running Lucent, so he didn’t give a crap. At the time of the breakup they made the package so attractive that about 1/3 of our center left, no one having trouble getting work. My department had far more money than people at this point, and was bring in chemists to do EE work.
One of Bell Labs marketing tricks was to confuse outsiders about what it did. Area 11 did research. I think there was a little bit of A11 at Holmdel, but very little. Most of it was in Murray Hill. A big chunk of Holmdel worked on transmission stuff, consumer products, and datasets. I got offered a job working on Dataphone II, and early modem, there in 1980.
Most of Bell Labs did product development - hardware design, software, test development, the whole bit. A11 was chartered to work 3 - 5 years out. My center had research money to work on manufacturing stuff 1 - 3 years out, and my group and I were paid for 50% research and 50% development.
The point is, someone looking at the big buildings shouldn’t get the impression everyone there was doing fundamental research. Doing technology transfer was just as hard there as anyplace else. A11 people were strongly rated on paper production. We worked with a guy who had a nice algorithm for protocol testing which we were able to adopt into an electronic test generator and even a product. He even did time on a trade show floor with us. This did not win him a lot of brownie points with his management.
And of course Penzias and Wilson were doing pure development when they made their discovery.
As mentioned, Microsoft and Google are still big on basic research. I knew a guy who used to do research at IBM and he told me 20 years ago that he could see the tunnel at the end of the light. One problem is that companies resist using the results of their research since it might undermine standard business. All our windowing interfaces grew out of work at Xerox, but the company was unable to capitalize on it. The founders of Google were in academic research at Stanford and understood the value of research. In fact, I heard Sergey Brin give a talk on his idea for a new search engine at Standford early in 1997. I only shoulda asked him if I could invest.:smack:
Yes, you should have. I think one of the professors at Stanford was an early investor in Google, investing about a million dollars but now is worth over a billion because of that early investment.
And I think a lot of corporate R&D has shifted to the universities, to places like the MIT Media Lab for pie-in-the-sky stuff.
It wasn’t all Product development, though. I visited people at Bell Labs Holmdel who were doing actual research, and I knew others who went to work there – sadly, shortly before the breakup.
Yes, IBM’s giant funding pot for research came from the massive markups on mainframes, since it was practically a captive monopoly market.
I recall that for the first AT machine (286) - rumour has it they were going to produce a 12MHz and the System/38 people saw it and said “You can’t ell that! It will destroy our business!” ($7,000 for a desktop PC vs. hundreds of thousands) IBM sold a 10MHz 286 while Compaq soon made a 12, then a 16MHz. IBM never really got back out in front on the PC market.
Similarly, yes, the Apple Lisa (and then Mac) was inspired by a Jobs visit to see the Xerox Parc effort, including the mouse. Look around today, and Xerox is just another copier company, they failed to capitalize when they had the monopoly. Apple is one of the biggest companies in the world by stock value.
When I joined a smaller multinational many years ago, engineers were treated as something special. They were encouraged to do basic process research in the field, publish internal papers that ocasionally made it to scientific journals, there was a long-term research facility, etc. then wall-street valuation techniques came along, head-count was “shrinked” based nf appropriate ratios from the big boys in the finance institutions, engineers simply became process technicians, and the research facility was shrunk because it cost money and had little short-term payoff, and eventually even after all this cost-cutting, they were still a sitting duck for an unfriendly take-over… meaning even less worry about research and even more cost-cutting to justify the buy-out price.
I suspect that a lot of North AMerican companies did something similar - research costs money, the premium-price gravy train was over, research is the first cost to be cut.
My concern is when governments start behaving in the same way. Afterall, the CEO of the UK is probably looking at a not too dissimilar term in office to the CEO of pfizer. I can see the motivations for curtailing free-wheeling research in a modern corporation (but not necessarily accept them - it would be completely wrong to attempt to apply free-market principles to the process of scientific discovery). So it falls to the government to supply longer term funding to support fundamental science.
What we’re seeing in the UK, though, is government incrementally trying to set the scientific agenda of what can and cannot be funded. We refer to something called the Haldane principle here, named after a UK politician, which just says that you need to trust scientists to decide how best to fund science. If you don’t do that then you end up with a planned scientific economy that actually doesn’t work that well - because who can predict the pathway to scientific breakthrough?
I used to go to Holmdel all the time. There was no research to speak of (except for maybe an MTS or two who got themselves transferred) in Indian Hill, Allentown, Denver or the Merrimack Valley Labs. Some in Holmdel, mostly in MH.
I don’t remember the exact counts but Bell Labs was 90% development and 10% research - though I suspect the average person would think it was he opposite. That’s the way the press releases ran.
Since I got funded by groups all over the company, I spent a lot of time visiting Labs and factories all over the country.
Let’s instead call it the Viagra Principle. The biggest moneymaker in recent pharmacological history was entirely an accident. People taking a test heart drug reported an unusual side effect.
The moral being, you explore everything because you have no idea where it will lead you in the end.