What Google et al are doing is a lot like the Bell Labs of old. Just as AT&T’s monopoly profits paid for a lot of research of uncertain short term direct value to Ma Bell, the Internet bigs have poured a lot of money into side projects that may not actually be profitable, but the core business prints so much money that investors aren’t inclined to ask too many questions.
In hindsight, we owe a lot to Bell Labs and Xerox PARC. Maybe in twenty years the effect of research projects at Google and Amazon will be just as significant.
Of course, as great as Bell Labs was in 1960 – 1980s, I don’t believe the rollout of cell phones would have been as great if not for the breakup of Ma Bell. (Especially since Ma Bell spun off the cell phone businesses to the local companies.) It does appear for the basis of discovering innovation, you need a tight Monopoly or Oligopoly (think pharma) but for consumer rollout you need a competitive market economy. That is best part of Al Gore internet web rollouts which did little to solidify the internet property rights and led to a lot of internet ‘Oklahoma land rush’ ideas. However, I find very weird how quickly Internet markets solidifies to Monopoly/Oligopoly models very quickly despite the companies not needing high capital expenditures. In terms of Google or Amazon, they really don’t protected monopolies and yet control a good portion of the market. (This is one reason I don’t understand VC money going/losing/burning into Uber….It seems like a market that would easily be added entrants here.)
In terms of alternative, I have always wondered what would have happened to Ma Bell if Reagan (especially)/Ford had won in 1976 and placed Judge Robert Bork on the case? Otherwise, Carter placed Judge Greene on the case who usually ruled against Monopolies and probably led the most successful deregulation of the era. (I always assumed Reagan administration did not react against the Ma Bell ruling because he was dealing with the 1982 recession and had bigger issues to face that year.)
Actually, it appears you need to weaken, as killing seems sadly beyond reach, the government regulators
“When AT&T wanted to start developing cellular in 1947, the FCC rejected the idea, believing that spectrum could be best used by other services that were not “in the nature of convenience or luxury.”… ”
http://www.aei.org/publication/regulation-vs-innovation-why-1950s-teenagers-didnt-have-cellphones/
The FCC was stupid because this cell phone stuff was probably 20 years out to begin with. (We tend to way under-estimate the length it takes most technology to roll out. The internet was invented in 1969.) Also I said it is best for a Monopoly/Oliogopoly for research and development (Ma Bell labs) but we need competitive markets for the best roll out. I seriously doubt if Ma Bell had a cell phone monopoly it would have roll out as fast or great. (Especially since the cell phone market is now crimping the local telephone service.)
The article saying 1950 teenagers carrying cell phones seems incredibly overstated as it took about 20 years (1983 – 2003ish) before teenagers started carry cell phones in our world anyway. I think most teenagers started carrying cell phones when providers rolled out big family plans in the early 2000s. Sort of a price discrimination model with high fixed cost/low marginal cost at the time.
This analogy is very strong. Running a telephone system at American scale was something no one had done before. Much of the stuff bell labs proved directly useful, all kinds of electronic parts used in networking today were developed at bell labs and put immediately into use.
Google’s work on AI is similar. They took natural language parsing to heights never seen, and we all use it everyday. If we get true general AI it will be because google really wanted it’s algorithms to understand the context of text on webpages. Google photos is great example of being able to search images because google developed the AI to understand what is happening in a picture and write a description, and then search the descriptions, and they are doing it at scale that is impossible for anyone else.
Just like signal amplifiers on phone wires, googles tech will eventually become a tool in the box of every engineer, and their monopoly wont’ let them own every use of these products.
It’s been said that the biggest key to successful R&D is simply this: consistent funding.
Alan Kay, one of the researchers as Xerox PARC, gave a a couple years ago to business leaders about successful R&D. He mentioned that PARC invented (among other things):
– personal computing
– the graphical user interface
– laserjet printing
– much of the foundation of the Internet (including Ethernet)
All with 25 researchers at the cost of about $12M (in today’s dollars) a year. Laserjet printing alone netted Xerox about 200 times that cost.
The main obstacles to successful R&D are:
– few companies have the discipline to invest research that could take 10+ years – instead they’re constantly driven by quarterly profits (the decline of Hewlett-Packard can largely be attributed to this)
– Few managers have the talent of hiring and managing people who are smarter and more creative than themselves
There are many things Google is developing that have not shown their main fruit.
Google Scholar opened the door to the open science movement and journals, for example. Google Books has yet to have its full impact, but one derivative was announced recently as the Venice Time Machine gets going…