Don’t answer until you’ve read the following

Stephen Addison

TODAY WE HAVE a number of companies—Google, Facebook, Amazon, Microsoft, and Apple—that are considered IT monopolies, and a lot of people think their activities should be restricted or even that they should be broken up. But whenever anybody suggests breaking up IT monopolies, I begin to think about the consequences, many of which aren’t the sorts of things I would like to see happening. There would be impacts on research and the nature of research and the amount of research that was being done; there would be impacts on national economic competitiveness; and there would certainly be impacts on national security. All of those impacts are worrisome.

If we look at some history, I think we’ll have a better sense of what we’re talking about. For many years now, research has basically taken place in three places—federal labs, universities, and industrial labs—with most of the research in universities actually being funded by the federal government. The research conducted in these three places is likely to be “directed”—that is, guided and controlled by the funding agencies.

But in the 1940s when Vannevar Bush was setting up the National Science Foundation, groundwork that led to the development of federal research funding after the Second World War, he emphasized and championed the idea that research ought to be “curiosity driven.” In other words, people would choose their own topics. Quite often that research might begin with no particular application in mind, but history has shown that lots of breakthroughs are made that way.

Industrial research in the U.S. basically started in two labs. One of those was the General Electric Laboratories, which were founded in 1900, and the other was AT&T’s Bell Labs, founded in 1925. Those labs were dominant in the 20th century. But today General Electric has fallen on hard times. It was de-listed by Dow Jones in 2018, and in a sign of how far it’s moved away from electricity and research, its dominant business at its de-listing was GE Finance, which caused its problems. Just this year, GE split up into three companies: GE Healthcare, GE Aerospace, and GE Verona, which focuses on energy. They’re not doing any of their old core businesses and they have no effective research lab these days. That was basically sunk by the capital crisis of 2008.

As for AT&T, it was broken up by the Justice Department. There’s an AT&T today, but it’s not the same company. It’s a successor company made up of the “Baby Bells” that were founded by the breakup. Bell Labs, which was the jewel in the old AT&T crown, still exists—in name only—as Nokia Bell Labs, but it’s not doing the sorts of research it did back then.

For most of the 20th century, Bell Labs was key to the scientific enterprise in this country. These labs did work that supported the development of companies, but they also funded lots of long-term basic research that had no immediate commercial application. While, as I mentioned, that research is generally called “curiosity driven,” a Bell Lab scientist named Andrew Odlyzko originated an even better term for it: “unfettered research.” That’s the way I like to think of it—as research not driven by research budgets, not driven by the need to submit reports, not driven by the need to publish, but research that basically was agreed on by colleagues and managers as worthwhile to pursue, but with no necessary outcome. Unfettered research probably was more common at Bell than anywhere else before or since.

And just look at what it gave us. It gave us information science. It gave us the semiconductor industry. Both of those things together have led to the computer revolutions, and that was all started in unfettered curiosity research at Bell. It also led to the development of many computer languages. It led to the development of the UNIX operating system. All of those things were developed at Bell, and it resulted in 10 Nobel Prizes, five Turing Awards—which are for computer science—plus five Draper Prizes, in engineering. They won award after award, extending from the Nobel Prize down to the Academy Awards and the Grammys. Hundreds of labs were set up that were modeled after Bell, including by Exxon, Standard Oil, Mobile, IBM, and Xerox.

***

THE BELL LABS of today is just a shadow of its former self. It doesn’t support unfettered research. Instead, it does research for a former telephone manufacturer in Nokia, because Nokia ultimately ended up owning the remnants of Bell.

But the IT monopolies, at least some of them, do the same sorts of things that Bell Labs did in its prime. All of them spend lots of money on research and development, some of which is more visible than others. But since the government seems to want to break up Google, let’s look at some of the things that have come about as a result of Google research. There’s their own Office Suite, which is made freely available. There’s Google Maps, which I think most people would agree is a great trade-off for Google’s being a monopoly. People can’t find their way anywhere without it. And don’t forget Google Drive, Google Translate, Google Earth, and the less familiar Google Colab, an environment that’s widely used in the data science industry. Plus there is Google’s effort in climate research, A.I. research, and medical research. In the year that wrapped up at the end of March 2024, Google spent $45.86 billion on research and development. Its revenue was a bit over $318 billion, so Google actually spent about 14.4 percent of its revenues on research and development, and it spends more and more money on R&D every year.

I don’t think the average person would benefit from a breakup of Google. I believe it would reduce our dominance of the Internet and would weaken us as a nation, because controlling the Internet and having a uniform system actually helps in projection of our national interests, including our economy and our security. Approximately 94 percent of all searches go through Google at this point, so if we dismantled Google there would be lots of other search engines, meaning a massive duplication of effort, and ultimately it would mean that the cost factor would go up. And who would be responsible for end-user security? As an additional drawback, the successor companies wouldn’t be able to spend $45 billion on research, and we would all suffer from it.

The current Google debate reminds me of when AT&T was broken up. I don’t think we’re better off because of AT&T’s demise—just look at the amount of your cell phone bill today. As for the companies that caused the split-up of AT&T, principally MCI and Sprint as I remember from those years, neither one exists anymore.

But I can well remember when the people from Sprint and MCI would call me up and ask what it would take for me to switch my long-distance service. I had a ready answer: “You have to develop an industrial lab,” I said. “And when scientists at that lab have won four or five Nobel prizes, come back and talk to me.”

————————————————–

Dr. Stephen Addison is Professor of Physics and Dean, College of Science and Engineering, University of Central Arkansas. He is also as Senior Member of IEEE, a member of the Arkansas Academy of Computing, and Secretary of the Arkansas Academy of Science.