How’s this for a tagline? “The History of Telecoms is at an end.” What’s in this story is a pretty solid analysis of the issues that the telecom industry faces. The basic thesis of the story is presented in a book that says that telecom is now an irreversible commodity, that we already have what we need, and that the experience that users are prepared to pay for can already be obtained.
There’s a lot of truth in this, much of which is presented in the reference above. I’ve pointed out much of it in my own blogs. We can deliver home video at 4K with current technology and prevailing service speeds. There is limited demand for 8K, and a lot of things would need to be updated to enable any market at all. Response times needed for most applications can be met even at broadband speeds lower than the regulatory minimums, and mobile devices constrain the capacity needed to deliver content outside the home.
What this boils down to is that we’ve adapted to the limits of best-effort Internet, and that adaptation has been very successful for both residential and business users. If we assume that nothing threatens our accommodations, then the thesis of the story is undoubtedly correct. Further, it seems that both telcos and their equipment vendors are resigned to, or even perhaps committed to, making sure that the accommodations are indeed successful.
What would our lives be like if we had hunkered down on telecom as it was in the early 1980s, though? What would business be like under that same assumption? There were plenty of telcos and vendors who were happy to do just that, or to support “evolutionary” changes like (dare we recall?) the Integrated Services Digital Network (ISDN) or Asynchronous Transfer Mode (ATM). Why didn’t we? The reason is that outside forces changed the demand paradigm, and that change created an opportunity for a shift in telecommunications service opportunity, technology, and vendors.
The “history of telecom” shows, then, that what keeps networking from falling into the commoditization sink is not advances in networking, but changes in what we want it to do. That means that the question right now is the question we could, perhaps should, have asked back in the ‘70s and ‘80s, which is not what breaks us out of the mess but who moves to bring the breakout real. Can telecom, meaning the operators or vendors, pull itself out, or must some deus ex machina intervene? If the former, how? If the latter, who?
The broadband age we’re in now was created, according to CERN, by the worldwide web and not the Internet. “Tim Berners-Lee, a British scientist, invented the World Wide Web (WWW) in 1989, while working at CERN. The Web was originally conceived and developed to meet the demand for automated information-sharing between scientists in universities and institutes around the world.” They’re right. The HTML concept, the web page and browser, were the core concepts that drove a demand-side opportunity. The success of the broader Internet came about because of three factors, and that then gave us today’s telecom.
The first factor is obvious; the web concept grew out of a community served by TCP/IP, and so the tools were transportable directly to that protocol framework. In addition, the concept of delivering hypertext material via the network would have to be implemented on anything else, where open implementations were available for the Internet.
The second factor was that, at the time, the only data network equipment available other then the Internet was proprietary to a host of vendors and fairly expensive. IBM’s System Network Architecture (SNA) had single devices (the “routers” of SNA) that cost perhaps five times as much as IP routers from vendors like Cisco, and router vendors quickly recognized the benefits of building adapters to carry other protocols.
This led to the final factor, which is that IP networks could be deployed as data services, which would serve both the consumer appetite for the web and business needs for application networks. This, of course, is exactly what happened; IP VPNs and the Internet could share a lot of elements, making the Internet the modern “dialtone.”
What these factors tell us is that telecom impact came along rather late, at the end of a fairly complex and perhaps uncertain chain of developments. It was a top-down process, and it succeeded where ISDN and ATM did not, simply because you can’t push spaghetti uphill. Demand, and opportunity, justify the investments in supply. Thus, telcos cannot, by offering “improved” broadband Internet connectivity, push developments that would transform telecom. Neither can their network vendors, by pushing products designed to make those “improvements” possible. 5G could never have saved telecom, nor will 6G.
So we have to look to the other “who” in the picture, and in doing that we have to assess how these three factors could be repeated. The key here is for a “who” candidate to be able to address that second factor, which means that some “application framework” (the first factor) is translated into an implementation that can spread and become an industry standard. The easiest way to make that happen would be to open-source the implementation of a framework. That implementation could then spread to create a community that would drive a service. But…and this is a critical “but”…if telcos are willing to do only connectivity, then the “service” has to validate some form of enhanced connectivity. That, I think, would have to mean that either it generated a need for raw bandwidth, or that it required something other than “best efforts”.
The “who” point comes down to the old “cui bono”, or “who benefits?” Who would benefit enough? I think the answer is that it would have to be a player close to the application side. Compute or software, in other words. Such a player could conceptualize a new application framework and offer an implementation. It would be easier for a hardware player (systems or chips) to do this because of the requirement to open up an implementation. So, I’d suggest one of the computer vendors (Dell, HPE, IBM) or a chip player (AMD, Broadcom, Intel, NVIDIA). These vendors also have an advantage over network vendors in being more aligned with application creation and hosting than with delivery, and new things demand asset creation.
Is this the end of telecom history, though? Obviously it is not, because absent a delivery model nothing much that gets created by the Internet or IT innovation process will ever reach users. But it likely is the end of the “easy” period of telecom history, when what we had to deliver was waiting for a mechanism. It’s the “what” that’s now needed, and I doubt that telcos or telco vendors can shed their connection-centric service bias in time to play in the coming age.
