On August 16, 2023 the Marconi Society is delighted to host three of the most impactful innovators of our time for an intimate, member-only event – 50 Years Ago and In the Future.
Vint Cerf, Internet Pioneer, Marty Cooper, Inventor of the cellphone and Federico Faggin, developer of the microprocessor, will share their stories, insights and predictions for the future in an open discussion with questions and answers.
Vint, Marty and Federico are kicking this event off by sharing a few of their ideas. Please join us to hear more.
What changes did you envision when you created your transformational innovation? What did you predict correctly and what have you been surprised about?
Vint Cerf: Bob Kahn and I expected that interconnecting computers in a network would lead to significant sharing of research results, including software and computational resources. That was the original motivation for the Arpanet experiment, including the demonstration that packet switching technology would prove to be useful for computer interconnection. We extended that concept through the Internet experiment that interconnected computers and packet networks with very different characteristics. The TCP/IP protocol suite overcame these differences to offer a more uniform interface for host computers on a network of networks.
I was not surprised by the flexibility of the TCP/IP protocol suite to accommodate very different kinds of packet networks. I was surprised, however, by the arrival of the World Wide Web (WWW), thanks to Sir Tim Berners-Lee’s HTTP protocol, and HTML encoding standard. This triggered an avalanche of new applications, which was amplified by the arrival of the smartphone (Apple iPhone) in 2007.
The avalanche of content that followed the introduction of the WWW gave rise to demand for indexing of its content. Alta Vista, Yahoo! and Google and others all became important search tools for finding relevant content in a sea of web pages.
As bandwidths increased, I was not surprised to see video conferencing become a major application – we had been experimenting with that as far back as the Arpanet period. The arrival of neural network machine learning leading to speech recognition, text to speech generation, language translation, and now Large Language Models proved to be another surprise. Early work with the single layer Perceptron neural network was disappointing, but multi-layer systems proved to be dramatically more powerful. These tools are becoming an integral part of the WWW applications we see today.
Marty Cooper: It wasn’t merely a prediction – I knew that eventually everyone would have a cell phone. I joked, “Someday, when a person is born, she will be assigned a phone number. If she didn’t answer the phone, she has died.” There are now more cell phones in the world than there are people. The cell phone has become an essential tool of productivity, an extension of its owner’s personality. And yet, a third of the world’s population still don’t have cell phones.
When the handheld cell phone was invented, the Internet did not exist. We did not have personal computers nor digital cameras, and the large-scale integrated circuit did not exist. In 1973, it would have been hard to imagine a “phone” that incorporated billions of transistors, the equivalent of a supercomputer, a powerful digital cameras and access to all the knowledge in the world.
Federico Faggin: The idea of eventually integrating a CPU into a single chip – what defines a microprocessor – had already been advanced in 1967 by Lee Boysel at Fairchild Semiconductor. However microprocessors were not expected to occur before the mid-Seventies because the current MOS metal-gate technology was not dense and fast enough to allow them. It was the unexpected invention and development of the MOS silicon gate technology, the project I successfully led at Fairchild in 1968, that speeded up the timetable to implement all the missing ingredients to make computers on a chip: microprocessors, dynamic RAMs (random access memory), non-volatile memories, and CCD image sensors.
When I joined Intel in April 1970, I led the chip design of all the microprocessors starting with the 4004 microprocessor. To achieve the appropriate speed and circuit density, the 4004 required two inventions of mine: the buried contact and the bootstrap load. Furthermore I had to develop the entire methodology for random logic designs with silicon gates that were also used for all the early Intel and Zilog microprocessors.
The Intel 4004 became the world’s first commercial microprocessor when it was first sold in March 1971. In early 1971, I took over the project that became the world’s first 8-bit commercial microprocessor in April 1972. By the same time I had architected the Intel 8080, and I later led its chip design. The 8080 was commercialized in March 1974. It became an instant success, showing that microprocessors could be broadly applied to many more applications than first-generation microprocessors.
In November 1974 I started Zilog, Inc., the very first company dedicated to microprocessors. At Zilog I architected and led the design of the Z80-CPU, a third-generation 8-bit/16-bit microprocessor with twice the speed of the 8080. Using microprocessors, personal computers began to replace minicomputers. However, the importance of these new chips was also to create novel applications and markets in which a microprocessor-based computer with appropriate software could replace special-purpose hardware solutions that were too costly, inflexible and took too long to develop. Most of these new applications were incongruous and impossible to imagine when the smallest conventional computers were power-hungry, cost thousands of dollars, and were larger than a shoebox. How could a computer fit inside the body, a toy, or a pocket?
I therefore expected to be surprised by most new applications since there is no limit to what can be done with tiny, powerful, and cheap computers. Nonetheless, there were many that I would not have imagined, for example, having a microchip inside an electric toothbrush! Why not when one can buy for less than $1 a general-purpose computer the size of a pea that dissipates less than 100 mW?
Looking ahead, what emerging technologies do you believe will have the most profound impact on society in the next 50 years?
Vint Cerf: For sure, large language models and other neural network applications are bound to have power impacts on daily life and work. The “Interplanetary Internet” is set to support manned and robotic space exploration and commercialization in the 2020’s and beyond. The Internet of Things (devices with computing and Internet communication capacity) will become increasingly common.
Less certain – but no less intriguing – will be vehicle-to-vehicle communication assisting self-driving vehicles in their work. Machine learning of all kinds may inform diagnostic work (e.g. medical, maintenance) and discovery (e.g. drugs, asteroid threats). New forms of detection and measurement will create a deeper understanding of how the universe works (think: dark matter, dark energy, strong nuclear forces ). Medical diagnosis and treatment will advance with AI/ML, small molecule therapies, new diagnostic tools, and personalized treatment.
Marty Cooper: When artificial intelligence (AI) is added to a cell phone, the phone will become a true extension of its owner. The AI will analyze the behavior, habits and preferences of the owner, and will either create or locate appropriate apps to relieve the owner of the routine aspects of his or her life. The ludicrous idea of sorting through 4 million apps to find the right one will disappear.
Further, the cell phone will evolve into a group of sensors located on or in its owner’s body, selected based upon the owner’s history and DNA. These sensors will monitor key metrics that will anticipate diseases and provide for their cure while they are still nascent. A few generations from now, the concept of disease will be obsolete
Federico Faggin: Artificial intelligence (AI) is the technology with the most potential applications. In the last 10 years AI has reached a level of performance and readiness to turn it into a technology with arguably even more widespread applications than the original microprocessors. Natural-language programming, self-driving vehicles, and robots of all types are just a few examples of embedded AI. Here again, I expect that human creativity will surprise all of us.
Quantum computers will also come of age, allowing us to go way beyond the speed of classical computers, especially when applied to the simulation of living organisms. Such systems, being both quantum and classical, are way too complex to be simulated classically, Semiconductor technology will reach maturity in 10-20 years and Moore’s Law will end. Only biological organisms have the promise to give us a quantum and classical technology that may top semiconductors in terms of integration, speed, and energy consumption.
Genetic manipulation will allow us to create new microorganisms and new forms of multicellular life with incredible potential for good and evil. The challenging ethical problems currently posed by the widespread adoption of AI are trifles compared with what is in store for us 20 years from now in conjunction with genetic manipulation. Only a planetary-scale regulation may avoid potential existential risks for mankind and for the ecosystem.
Collaboration and interdisciplinary efforts have been instrumental in many technological advancements. In fact, this is why the Marconi Society focuses at the intersection of technology, policy and digital inclusion to support digital equity. What disciplines do you think need to come together to solve the issue of income inequality?
Vint Cerf: We need better evaluation of strengths (think Gallup StrengthFinder on steroids), as well as tools that augment human capacity to analyze large data collections and hypothesize new theories. We will need a deeper understanding of Machine Learning to use these tools more reliably. Advances in spectrum sharing and efficient utilization will be key to ensuring that everyone has access to the connectivity and speeds that they need..
Marty Cooper: Advancements and universality of education is the key to solving many of the problems humanity faces today!
Full-time connectivity is transforming our antiquated educational process from a classroom/lecture process to full time learning based upon a student’s continuous access to all the knowledge in the world. The roles of teachers and students are already changing dramatically and will be revolutionized in coming generations. The idea of a teacher lecturing to students who have full-time access to all the world’s knowledge is ludicrous. Teachers need to focus on how their students can discriminate between real and fake information, and how to apply the vast amount of information to real world problems.
In order to achieve the potential of a connected society, radio-frequency spectrum must be available. Our politicians and regulators have treated radio spectrum as a scarce resource. Yet, there has never been a scarcity of spectrum. Technology has always provided the capacity to accommodate the burgeoning applications in areas like communications, national defense, and global positioning. Further, technology already exists that will provide additional capacity for at least another 50 years.
Federico Faggin: Science, philosophy, and spirituality need to converge into a new worldview that replaces the materialistic and reductionist perspective that currently dominates society.
I strongly believe that it will be impossible to solve the many problems we are now facing if we think that we are perishable machines guided by the “survival of the fittest” doctrine. This is the same principle that fostered climate change, the proliferation of nuclear arms, AI-based cybercrime, as well as the income inequality mentioned. This same egotistical doctrine is also behind the endless series of wars that have affected humanity since recorded history.
I believe that humanity may overcome the challenging times in front of us only when science finally recognizes the existence of consciousness and free will as properties that cannot exist in deterministic systems.
Only strong cooperation can save us, not the life-and-death competition inherent in the “survival of the fittest” doctrine. We need to recognize that in the connected world, we live in, the good and the evil we do unto others will inevitably return to their doers.
I believe this is the only option left. We are indeed all living together on a tiny planet sailing in the unfathomable space of the universe. Only by working together with respect and love for each other can we overcome the serious problems we have created for ourselves.
