Grant Lenahan
Appledore Research Group had the opportunity to attend the NOKIA Bell Labs seminar honoring the 100th anniversary of Claude Shannon’s birth. Shannon’s work on information theory has been seminal in our industry, building the foundation for digital transmission, wireless transmission, and has been applied to myriad other problems. He is truly a giant among giants.
Routing list: Shannon, Bode, Nyquist and Black WOW!
While the event was intended to honor Shannon, it clearly also signaled the importance that the “new” NOKIA places on Bell Labs, its heritage, and the role of innovation in the industry. It included lectures by some of the brightest minds in technology over the past decades – including Irwin Jacobs or Qualcomm, Bob Metcalf of 3COM (the “Com family of companies” joked Metcalf), Eric Schmidt of Google and Henry Markram of the Brain Project (Switzerland), plus many more panelists and demonstrations.
ARG took away several clear messages that bear repeating. Metcalf, Jacobs and Schmidt all emphasized the need to tightly couple science, engineering and commercialization – whether traditional product design and manufacturing, or a more “cloud age” version as in Alphabet. Unsaid, of course, is how much of the work done concurrent to Metcalf’s at Xerox PARC was commercialized by a different company (Apple, and then later Microsoft) due to the weak geographic and organizational ties that existed between Xerox and its Silicon Valley outpost.
Technology History: It does repeat itself
Bell Labs appears to be claiming that it will help lead NOKIA to commercialize its intellectual property, which is still substantial. Later in the day, we took tours to see how Bell Labs research is being applied to next-generation wireless, access, and cloud systems, as well as to more general computing and cognitive capabilities. We wish them well, not just for NOKIA, but also for the good of our industry as it wrestles with more and more competition from adjacent industries.
Another of the messages throughout the event has been the progress of understanding logic and machine learning – from Shannon’s learning, mechanical mouse, to Markham’s work understanding the structure and operation of human (and other similar) brains. His work is truly amazing, and indicates that unlike most computer architectures, the brain operates far less deterministically. Rather, Markham’s team has shown that the brain has tow characteristics that I characterize as “Darwinian”. First, it has a fairly modest number of base component neuron-synapse configurations that reconfigure constantly to solve different problems – it is NOT hard-wired region by region (as often stated). Second, various regions – as they reorganize – compete on methods of understanding reality, and one of the many wins out (survival of the fittest). Presumably this success is recorded and we refer to that as learning, but I am paraphrasing a vastly longer and very complex (and excellent) presentation.
Markham’s Findings
Little of today’s information provided direct, tactical input to the research that ARG publishes every month, and yet it may be all the more important for it. Today was valuable to me because it reminded me of first principles, and to think differently. It reminds us that tomorrow need not look like today; that we can make step functions; and that even in day-to-day operations we need to look at the leading edge of technology – not so much for what it can do today, but for where it may take us.
Appledore Research Group applauds NOKIA Bell Labs for reminding us all of this, thanks them for including us, and urges them to work hard to take worthy ideas out of the lab and into the network, the datacenter, the Telco Cloud we strive to build.
Grant Lenahan
Partner and Principal Analyst,
Appledore Research Group