Behavioral Consensus in Autonomy

A few months ago, I wrote a short manifesto about these 19th-century predictions, predictions of the future, and of our current:

The future of the 19th century was a romanticized projection of flying cars and communication technology upon common lens community. It is easy to laugh at the oddities of world collision and to overlook the energy of a world on the edge of flight and community building. Yet, it is hard to understand that today’s romanticized fringes may never become the driving forces of tomorrow — while the overlooked do. Manifesto: build for the human condition, not what we perceive to replace it. The future will mirror what has existed, not what could exist.

As a designer and a technologist, much of my life is spent thinking about the future, and about how I can build that future. My friends are all tinkering away on grandiose visions, wishing to radically shift the status quo and to invent the next big thing. Silicon Valley was built upon the mantra of doing what you cannot, and failing for success. Yet, often, I see my friends, and my industry, shifting from their true purpose. They design and build for a future that they wish to exist, not the future that will exist.

The stunning paintings of a 19th-century portrayal represent a fascinating distortion; one that is both correct, but amusingly wrong. Helicopters and aircraft permeate our world, beams of light deliver us messages from the sky, drones deliver trinkets and snacks, tweets and social media compress knowledge into our eyes in way we could never have imagined, and incredible connectedness has grown inescapable. But, the predictions in the above paintings get something very wrong: they ignore our humanity. The technology portrayed is shockingly accurate of today’s cutting edge, and tomorrow’s ubiquity. Yet, the technology is crammed into a world not built for their interaction, and a world blind of their value. Aircraft, social media, and connectivity don’t replace humanity: they improve and enable it.

The world that will exist tomorrow is a world that is built for people — with people. Technology is an enabler, not a replacement. Aircraft and connectivity have made the world small, but only because they live outside of it — planes take to the sky, and connectivity takes to conductivity. Our world is built around interaction and community, not the escape of it. Yet, these portrayals continue to shape our world. Since the 14th-century, visionaries and inventors have flocked to the ideological precedent envisioned by Leonardo da Vinci. A world where people take to the skies like birds, not as a luxury, but as a way of life. Personal vertical take-off and landing aircraft, as they are now known, have been thought to permeate our air and cast our world with noise. This is the future that Silicon Valley is building right now. Uber, Alphabet, and Apple all wish to take to the skies. First, the next-leap must be on our streets, and maybe under them, before VTOL is plausible, they envision.

For centuries, this has become our obsession, our grandiose — our obvious future. But, how could one conception shape our perception? Fiction. Humans have succeeded because of our enthrallment of imagination, because of our understanding and projection of what is, what is not, and what could be. Leonardo da Vinci, an artist, envisioned a future — and had begun to engineer it. His art, and his visual portrayals, spread, and ignited. This force, as it has come to be known, is called the Star Trek effect. That is, the fictional Star Trek universe of the 1960s became a cultural phenomena, and as a result, its art, its projection, became a road map. The youth that watched Star Trek’s characters wirelessly talk to each other through a Communicator are the generation that built the cellphone. The future is not possible until it is envisioned, and with vision, comes possibility.

The world that feels so omnipresent and ever-possible today is that of autonomous vehicles. Self-driving automobiles, of course, are not a new prediction — they are a centurial obsession. While General Motors was shockingly accurate in their above portrayal of the problems we will face — the transition from safety-riddled transportation to socialization on-the-go—they have overlooked the human factor. General Motors’ vision was built upon 1950’s Manifest Destiny, one that is under millennial siege. John Zimmer’s Third Transportation Revolution is underway, and our cities are evolving to the future that should have been, the one where cities are built around people, not cars. The overlooked future will not be one of massive skyline highways, but instead of the death of vehicle ownership, rampant sharability, and greenery-doused streets.

The same could be said about the atompunky vision of flying cars that will permeate our atmosphere. The technology will exist, and our vision will have invented it, but its purpose will always emphasize our humanity, not the dystopia we envision. Human-as-a-requirement is a trend that is largely overlooked, but is something that differentiates success and failure. We are heading towards a future where vehicles become autonomous, not for luxury, but for way of life. Those new vehicles will need to fit into our ever-changing and ever-human world, and they will need to work well alongside us. We are on the cusp of machine intelligence that will see more, think faster, and execute better than us — yet, we are largely avoiding the problem of their behavior.

As of June 15, 2017, the Department of Motor Vehicles in California has issued Autonomous Vehicle Testing Permits to 34 private entities seeking to develop and test anonymous vehicle technology in the State of California. Those 34 companies are racing towards the vision of autonomous vehicles, and permeating the planet with the world-altering technology. Few, however, are designing in the open. Lawsuits and trade disputes are the new norm, and consensus is little. Silicon Valley is designing an autonomous future where vehicles and machines act like humans to replace the need for humans, but, engineers and designers are forgetting about the human factor. Operating a vehicle today is standardized: we, as humans, obey the rules of the road. We read hundreds of signs, steer within lines, communicate with those around us, and obey precaution.

However, the core reliance of driving is social behavior, because we are inherently social. We react in favor of us, preventing harm to us, and those around us. Drivers protect themselves first — a preservation instinct. Autonomous technology presents a future where the overwhelming majority of accidents and road errors never occur — but some will, sadly, still occur. It is the worst-case behavior that we must design, and must embrace in the open. Today’s self-driving cars are black boxes, each built around proprietary data, world understanding, and behavior. Tomorrow’s must understand the new rules of the road, a standardized consensus, one that is designed for, and by, humans.

Behavioral consensus is the greatest problem we face today, and it is a problem that will be solved by the many, not the few. Autonomous technology represents a new body that is more superior in senses than our own, but is naive in decision making. Decision-making skills will be built atop of autonomous technology, not within it. In the same way that the Institute of Electrical and Electronics Engineers, the Internet Corporation for Assigned Names and Numbers, and the Unicode Consortium have shaped computers and the internet, a new consortium and consensus must be brought to autonomous behavior. The Society of Automotive Engineers has standardized vehicles, and the National Highway Traffic Safety Administration has standardized roads — a new cross-proprietary, cross-border consortium must be formed to shape consensus in autonomous behavior.

As with roads, laws, and obedience, our world is shaped by agreement and predictability — choice is few and far between. For autonomous technology to truly permeate our future in the ways we envision, it must grow human — it must work with humans, and by human consensus. We are on the cusp of a new era of behavioral standardization, one that will not look far different from the internet. In the same way that Google’s Search has grown to understand consensus through repeatability, and Facebook’s M assistant has grown understanding of helpful interjection through usage, a new behavioral network will be formed, one in which use-case and consensus is oracle. Humans and machines are partnering, and they must be designed for the predictabilities of tomorrow.

Just remember: build for the human condition, not what we perceive to replace it. The future will mirror what has existed, not what could exist.