Innovation Physics: Part 2

By Harry R. Weller and Andrew A. Schoen of New Enterprise Associates

Note: This post is continued from Innovation Physics: Part 1. The technological “abilities” described in this post are emergent phenomena of the “tools” described in the prior piece. If you haven’t read Part 1, it can be found here.

This post is dedicated to Harry R. Weller. We worked together in framing it and writing it for over two-and-a-half years. The contents are largely his vision, which informed his investment decisions for over a decade. I learned an immeasurable amount from him through the countless discussions and iterations leading to this post. Though our conversations were always winding, those who knew Harry knew that he could always see around corners.

Emergent Phenomena

In philosophy, art, science and systems theory, emergence is the process whereby larger entities, patterns, and regularities arise through interactions among smaller or simpler entities that themselves do not exhibit such properties.

As mentioned at the outset, Innovation Physics is a narrative of tools and abilities. The tools discussed in Part 1 constitute the technological underpinnings that together form a highly complex system which gives rise to novel, unexpected abilities. These abilities are emergent properties in that they stem unpredictably from the collective properties of the constituent parts, properties that the individual parts do not themselves exhibit.

In some theories of particle physics, even such basic structures as mass, space, and time are viewed as emergent phenomena, arising from more fundamental concepts such as the Higgs boson or strings. In many interpretations of quantum mechanics, the perception of a deterministic reality, in which all objects have a definite position and momentum is actually an emergent phenomenon, with the true state of matter more aptly described by a wave function. Classical laws of physics, matter, cells, organisms, populations, species, ecosystems, societies, and economies are all emergent phenomena stemming from the interactions of complex systems.

New tools often give rise to unexpected emergent abilities, and these abilities often reach far beyond the tools' originally envisioned use cases.

Consider that the invention of the hand ax may have led indirectly to the creation of humanity’s first artwork. The hand ax unlocked precious time and calories that could be put towards pursuits other than acquiring food and surviving, thus enabling early humans to expand their mental horizons.

Today, the rise of computing power and its quickly multiplying applications are fundamentally based on emergence. Information theory is the field that describes these principles at a high level. In short, all modern computers use bits. Bits are binary and can exist in only two states, which we represent with a 0 or a 1. By stringing together series of bits and performing basic transformations to their states (via a process called logical gate operations), computers can do everything from calculating simple arithmetic and word processing, to running video games and simulating reality. Just as with the hand ax, computers can accomplish much more than their originally envisioned use, which was primarily the automation of calculations. Today they enhance nearly every aspect of human life.

Through the process of emergence, the tools described in Part 1, the Four C’s—computing, consumption, connectivity, and componentization—give rise to emergent abilities, three of which we identify here: antifriction, virality and judgment.

The Abilities

Now that we have briefly explored key underlying trends in the development of digital tools, let us turn to the abilities these developments enable. Three emergent forces, in particular, arise from these technological underpinnings: they are antifriction, virality, and judgment.

1. Antifriction describes the unhindered flow of information. Whereas friction is the force resisting objects sliding against each other, antifriction is the slippery and easily transmissible nature of digital information in today’s internet connected world.

There are both positive and negative ramifications of this transmissibility. Control over information is moving from centralized to decentralized locations. Antifriction accomplishes the spread of influence, a democratizing power unimaginable just a few decades ago.

For example, whereas industry titans once held all the answers for manufacturing and distributing products at scale, in today’s digital environment, goods, services and content can go from creator to market at the speed of the internet.

A creator can now use Kickstarter to raise funds and bring a product to market rapidly, without special access or influence. Faster yet, they can print their product on a 3D printer and be ready for shipment in the time it would have taken to communicate with a factory. Or, even faster, they can provide the goods or services virtually.

Antifriction enables technology startups to scale and reach their market efficiently. A quick glance at the music industry illustrates the vast increase in the availability of music, the speed of delivery, and the ability for anyone to publish their creations. The publication industry similarly democratized; the editor and publishing house are no longer needed to shape, ordain and distribute content.

Antifriction enables creators and service providers to reach their audiences and customers directly. Therefore, the drivers of lock-in—defined as control, power and influence over markets—are no longer economies of scale and supply chains; they're network effects and data.

2. Virality is the tendency for select units of information to replicate and propagate rapidly and broadly.

Virality plays an important role in the dissemination of all types of information, from news to research to entertainment to products and services. As digital technology increasingly permeates the modern world, virality’s role can only amplify in importance. Ideas and products will propagate increasingly faster and wider.

This propagation of information is not dissimilar to virality in natural systems. Units of content move through the digital ecosystem with varying degrees of success based on survivability and replicability. As in complex natural systems, it is not always clear which features of content will lead to survival and propagation. Vectors and mechanisms by which successful content moves and interacts can be studied in situ, and while it is possible to engineer certain viral properties into units of content, it is difficult to do so with consistent control over a specified outcome.

The rate of adoption for new technology products is increasing over time due, in part, to virality. This phenomenon is neatly encapsulated in the shortening timespans of major consumer applications reaching their first 100 million users. In the summary table below (compiled for this project by the Cornell Venture Capital Club), companies in each category are ordered chronologically by launch date:

3. Judgment is the ability of machines to make decisions on their own. Computing at its core allows for the automation of mathematical processing. As processing power, methodologies, and computing infrastructure improve, the scale, complexity, and precision with which information can be turned into insights also improve. This progress can be summarized as the ability of programs to take diffuse information and make a judgment. The ability to make informed decisions based on a collection of information is the basis for the evolution of artificial intelligence. In this context, the term judgment more precisely describes the ability of programs to make decisions and distinctions about data. What is remarkable about the force of judgment is that it is now an ability which can be automated.

The Netflix algorithm is a good example of the ability of judgment: “If you liked that film, you will probably like this one.” The ability to distinguish and segment target audiences and individual preferences from diffuse bits of data streamlines content discovery for the user and makes marketing and content distribution dramatically more efficient for the supplier.

Advertisements, entertainment content, news and political information, search results, and even the prices of goods and services can be pre-optimized and delivered to select target audiences for zero marginal cost.

This ability of programs to exercise judgment can be replicated millions of times using vast populations of internet users with increasingly vast collections of data containing previous user behaviors. Your behaviors on the internet influence the content you will see on the internet in the future. These predictive processes are constantly refined, redirected, and reinforced. One has only to understand the automation of judgment to realize that our experience of the world through the internet is evolving dramatically as the internet begins to know us as well as we do. The force of judgment underpins the phenomenon that “you no longer surf the web; the web now surfs you.”

Advances in artificial intelligence, particularly in machine learning and various other optimization techniques, are driving the ability of computers to simulate and exceed the capabilities of human judgement.

The above video visually illustrates a genetic algorithm optimization technique. This simulation utilizes a basic physics engine and a simple simulated 2D topology to evolve vehicle designs. It first assembles components (shapes like circles and squares) at random and tests the randomly spawned vehicle designs. It then utilizes the performance (distance traveled) of these simulated vehicles as the basis for selecting vehicle designs to be passed on to the next generation (plus a degree of random mutation). Through this process, the algorithm automatically evolves vehicles and optimizes their performance.

In addition to improvements in code, radical improvements in the computing substrate (e.g., hardware) could massively magnify the ability of digital technology to simulate judgment.

Conclusion

Predicting the emergent effects of the contemporary wave of innovation is not an easy feat, but it is among the chief aims of a venture capitalist. The advances in technology are profound. In essence, digital technology is becoming ubiquitous and resembles a new fundamental medium—a new oxygen—surrounding modern human life.

Four digital tools underpin the contemporary digital technology revolution: computing at relatively low cost, consumption of applications through powerful user interfaces and devices, connectivity provided by better and broader networks, and componentization allowing the efficient scale out and replication of new ideas. In the developed world, this increasingly ubiquitous set of tools is more aptly characterized as a medium surrounding modern human life than as a discrete and distinct set of experiences and accouterments. Like the air we breathe, the applications we interact with on a continuous basis are a crucial but increasingly unnoticed feature of existence.

The collective interactions of these four tools enable at least three emergent abilities: antifriction, virality, and judgment. 

Antifriction allows for efficient decentralization of creation, production, and distribution. Virality allows for rapid transmission of product, content, and ideas at increasing rates. Judgment allows for the automation of optimization and decision-making.

At the essential level, tools are an extension of the self that enables an individual to better accomplish a task. Following a trajectory similar to the “hierarchy of needs” described by Abraham Maslow in his famous 1943 paper “A Theory of Human Motivation,” tools have evolved from an extension of the base self (e.g., a hand ax to meet and provide for survival-related needs such as food, shelter, and safety) to an extension of the higher self (e.g., digital technology empowering communication, invention, democratization, education, science, creativity and self-expression), ultimately towards meeting and providing for self-actualization.

Given this upward movement in the ability of tools to satisfy the hierarchy human needs, it is interesting to remember that the first modern computer was built for survival. Alan Turing’s automatic computation machine helped break the Enigma Code and saved lives during World War II. Now computers and the applications that run on them help tackle higher order needs — business, politics, scientific research, creativity, interpersonal communications, and relationships. There will always be higher order needs, and the indefatigable march of recursive technological progress will continue to enable humanity to unlock and meet those needs, moving to ever higher rungs of Maslow’s hierarchy.

This article originally appeared as a post on the NEA Blog.