Q&A | Steve Wolff: A Father of the Internet

Recently, Science 2034 had the opportunity to hear from Steve Wolff, one of the fathers of the Internet and Principal Scientist at Internet2, which provides the computing backbone for the research and higher education community in the United States. At the National Science Foundation in the 1980s and early 1990s, Wolff led the development of NSFNET, the precursor to the Internet we know today. In this Q&A, he looks back on the milestone events that influenced the development of the Internet and shares his thoughts on what advances in science and technology hold for the Internet of the future.


Looking back, what were some of the notable events that shaped today’s Internet?

STEVE WOLFF: From the 1970s to the early 1990s, many decisions and events – some more memorable than others – formed the Internet we have today. 

There was, for example, the very significant decision by the Defense Advanced Research Project Agency (DARPA) to fund the incorporation of the Transmission Control Protocol/Internet Protocol (TCP/IP) into an advanced Unix operating system being developed at the University of California, Berkeley. This was followed by the milestone agreement between the National Science Foundation (NSF) and DARPA in the late 1970s to allow Computer Science Network (CSNET) traffic to travel on ARPANET circuits, and then by NSF’s decision to specify DARPA’s TCP/IP protocols for NSFNET. These three events effectively united the ARPA and NSF network research, development, and use communities and provided for a unified forward path under the guidance of the Internet Engineering Task Force (IETF), which remains the primary Internet standards body today. 

A fourth event, passage of the Scientific and Advanced  Technology Act of 1992, which sanctioned commercial traffic on the NSFNET backbone, ushered in a flood of creative activity and traffic nationwide. In this sense the Internet could be said to have been created by the U.S. Congress. 

Finally, two other events, the first virus – the Morris worm of 1988 – and the first spam, the infamous Cantor-Siegel green card incident of 1994, were not widely appreciated at the time. Their significance, however, is understood only too well today.

What are the forces that will shape the Internet of the future?

STEVE WOLFF: While there will be many inputs, two trends will have a significant impact on the Internet and computing capabilities of the future:

The evolution of the Internet as we will know it in 2034 will be influenced at least in part by a global increase in data transmission speeds.  The speed of NSFNET and its successors roughly has kept pace with Moore’s Law, which holds that computer processing power will double every two years. If the trend continues, the Internet2 Network of 2034, for example, will be running at 200 terabits per second – a rate that could transmit the entire collection of the Library of Congress (books, videos, photos, music) in 20 minutes.

The two decades since 1994 saw the speed of the fastest supercomputer in the world rise from 170 Gigaflops – 170 billion calculations per second — (the Fujitsu numerical wind tunnel) to 34 Petaflops (quadrillions, China’s Tianhe-2).  Further increases have been slowed by the need to moderate the machines’ thirst for electric power.  However, in the next 20 years, we will see a new class of computers that use cognitive chips that realize high speeds at very low power levels.  In 2014, IBM announced a revolutionary new architecture: a neurosynaptic chip with 5.4 billion transistors that is organized like a human brain with a million neurons and 256 million synapses.  By 2034, the most powerful computers in the world will use a combination of conventional and neurosynaptic architectures, the “flop” as a measure of computer power will have fallen into disuse, and a new generation of network communication protocols will have been developed to optimize communication among these hybrid machines.  Today used primarily for simple pattern-recognition tasks, by 2034 the ability of these hybrid neurosynaptic computers to exercise autonomic reasoning power on observed data will lead to a fifth paradigm of scientific research (i.e., the ability to deduce causation vs. the fourth paradigm’s correlation).

How will this impact education?

STEVE WOLFF: The higher education system will have been transformed by the Internet and by another 20 years of research in learning and cognition.  By 2034, 3D and immersive technologies will have revolutionized skill acquisition in the trades and professions, and immersive visualization will bring new modes of understanding to complex and even abstract problems in mathematics, engineering, and the sciences.  But the Teaching Associate – Research Associate – Postdoctoral Fellow apprenticeship system for entry into the learned professions will continue to flourish in the major research universities. 

At the same time, students will have become even more nomadic, grazing not only among university campuses but also at one-of-a-kind facilities such as the Demonstration (fusion) Power Plants (DEMO) follow-on to the International Tokamak Experimental Reactor (ITER), scheduled to begin operation in the early 2030s; the Very Large Hadron Collider (VLHC), now being discussed as a possible successor of the Large Hadron Collider (LHC); a now-hypothetical Ten Square Kilometer Array, or XSKA, incorporating and superseding the Square Kilometer Array (SKA), the world’s largest radio telescope; and other multinational experiments.

So what will the Internet of 2034 bring?

STEVE WOLFF: In 2034, we will all live in a soup of electromagnetic babel, the incessant chatter of the Internet of All Things, resulting from the many low-power devices in our environment and on our persons that will require no batteries, but rather will be “powered by the aether.”

But guessing the future, it seems, is chancier now than ever. William Gibson, author of the dystopic novel Neuromancer put it:  “[T]hat truth-is-stranger-than-fiction factor keeps getting jacked up on us on a fairly regular, maybe even exponential, basis. I think that’s something peculiar to our time. I don’t think our grandparents had to live with that.”

The Early Internet.
The research and education community played a seminal role in the creation and growth of the modern Internet and the applications that have placed it among the most transformative technologies in modern times.  In the United States, the fledgling Internet initially was nourished by the academic, military, and industrial research community, funded by the Defense Department’s Advanced Research Projects Agency (ARPA).

Later, when the freely available and open source UNIX computer operating system infused academic computer science departments, the USENET network spread rapidly throughout academe. First conceived and implemented at Duke University and the University of North Carolina at Chapel Hill in 1980, USENET was based on the simple communication protocol Unix-to-Unix Copy Protocol (UUCP) and was more akin to today’s bulletin boards or Internet forums.


In 1981, USENET was joined by the Computer Science Network (CSNET), funded by the National Science Foundation (NSF), and a grass roots network, BITNET, among academic mainframe computers.

Computer networking overwhelmingly became an enterprise of the higher education community when, starting in 1985, NSF funded NSFNET, a national backbone and family of regional networks created to interconnect the nation’s colleges and universities (based on recommendations in the Lax report of 1982).


U.S. R&E Backbone Network Speed

If you want to make sure that we realize the benefits from this and other scientific advances, let your members of Congress know that federal funding for science matters and should be made a higher national priority. The Science Coalition’s Science Matters toolkit can help you get started.
Categories Technology