Much of the technology required for a more advanced state is already here in the initial phases of the Internet of Things. The substantive difference between the Internet of Things and the Social Network of Things is actually more about the need for us to change the way we think about and use networked systems: moving away from thinking in terms of objects and humans and instead thinking of the world in terms of agency and cyborgs. These conceptual shifts are deliberately messy constructions.

When we have fully made this transition in our thinking, we will have reached the Social Network of Things (SNT). This next paradigm is partially explored by Bonchek and Choudary, but deepens in impact when you apply it beyond the conception of social products (useful, as the business portion of a larger trend) and platforms (again, a business-centric view). Both Bonchek and Choudary allude to (and understand) the larger shifts: that as we progress technologically it is less and less possible to decouple humans from devices and the shared purposes which bring them together.

The SNT touches on many of the same technologies which begin to emerge in the IoT.

When the Social Network of Things comes into full being, platforms will be fully interoperable, sharing data and computing resources among many devices and purposes seamlessly.

The SNT touches on many of the same technologies which begin to emerge in the IoT, like the cloud, machine learning and platforms—but it also begins to explore the implications of those technologies. Terms like Little Data™, the data surrounding individual users, will emerge alongside Big Data. Issues around robotics and new applications thereof—like robotic architecture which changes the fundamental shapes of buildings in response to their occupants or the weather, or to the concept of driverless cars—will bring about new functional and ethical issues. For example, how do we manage the negotiation and mitigation of risk between devices, such as two driverless cars coordinating with each other to avoid an accident?

At this point in the progression toward the SNT, systems and data have become interoperable and leverage rich APIs. Purposes for devices become highly shared, much as our smartphones have, so that our cars display our calendars, share processing power and network connectivity with a connected phone, serve as sensors and even power storage units on larger utility and transportation grids.

Design by our engineers and developers is premised in responsiveness—the ability to make adaptive, resilient platforms and devices which can change as their circumstances and users evolve. Information in this phase, unlike the simple tools at the beginning of the IoT Age, will take real-time information about the present and predict the future.

In order for our devices and networks to provide predictive information about the present, the transition to the SNT will occur on platforms which require interoperability, strong machine learning, and robust network bandwidth. Platform thinking will have become a de facto business strategy and technology doctrine. By today’s standards, the flow of data and evolution of devices will look more organized and reliable, but the overall technology landscape may look overwhelmingly chaotic and messy, because a very high rate of change will mean that we have as much culture work to do as coding.

The challenge is that the digital divide—the gulf between not just the haves and have-nots, but the knows and know-nots—will widen. Those with access to the right technology, bandwidth and tools will have exponentially bigger capabilities in business, health and other critical domains. Disparity will start to appear as much or more along educational and bandwidth-access lines than national identities, as globalization and urbanization play out.

At the same time, machines will be learning on their own, increasingly automating the tweaks to their algorithms. All of this will result in entirely new opportunities for value, new markets for services, products and platforms, and the ability for users to move to co-creators of their online and offline lives.

Tensions of the SNT

  • Fairness and equality

  • Safety conflicts

  • Legacy system retirement

  • Political and cultural backlash

Roles Necessitated by the SNT

  • AI Developer

  • Regulatory liaison

  • Cyborg Anthropologist

  • Machine teacher

Characteristics of the SNT

  • Messy and chaotic

  • Very high rate of change

  • As much culture work as code

  • Digital divide is enormous

  • Machines begin to learn on their own

  • Entirely new value opportunities and markets

  • Interoperability required for success

  • Mature platforms

  • Enormous security issues

  • Predictive information about the present

The Automated Car + Pedestrian Example: Why a Social Network of Things Matters

[Video pending upload and not yet active] This video shows a pedestrian in London crossing active traffic and how limited the options for conventional drivers are, posing questions about what is possible for the driverless or augmented-driver car. 

In the automotive example of the Social Network of Things, devices have multiple shared views of the world around them based on shared purposes like preventing collisions or easing traffic, enabled by pervasive network access and robust platforms for real-time machine learning.

In a real-life example where the SNT could have made a huge impact, a pedestrian crosses a busy street without a crossing signal. The first vehicle to pass by him taps its brakes quickly and then chooses to keep driving. The pedestrian runs between this car and another van behind it, which doesn’t even touch its brakes. In the world of automated vehicles, we have the capacity to ensure his safety and that of the drivers through the transmission of data between multiple devices and people instantaneously. For example, the humans here couldn’t convey from one vehicle to another what they had seen, or how they made the judgement call they did—but an automated vehicle could.

How do we design systems which can tap into the full potential of a social network of things—leveraging sidewalk sensors, the on-board systems of both conventional and driverless cars, traffic cameras and everything else available in the digital world—to augment human capacity?

Even more complex is the possibility of having these devices not just inform each other, but negotiate an outcome. Let us suppose that two driverless vehicles identified an imminent crash. Without being able to communicate with each other, as is currently the case with most driverless technology, the vehicles have to guess about what the other vehicle is going to do and use their on-board information to try to avoid or at least lessen an impact. If those cars could coordinate their decisions, they could synchronize their trajectory to more reliably lessen or avoid an impact altogether.

The US Department of Transportation and many other regulatory and industry agencies are spending energy and resources to get ahead of the coming complexity of such systems. It’s just in time, too—as evidenced in the 2013 Wired article “The Ethics of Saving Lives With Autonomous Cars Is Far Murkier Than You Think.

Life in the age of the Social Network of Things—at least, properly implemented—is potentially safer and easier. At this stage, we are doing something which has never been done before: negotiating between devices without passing decisions through humans first. In the world of mobility, we are attending to multi-modal transit, major safety issues and even the possibility of ultra-light vehicles which rely on intelligence, rather than mass, for safety. Some cars will be fully automated, while others will be augmented with simple systems like accident avoidance and traffic routing, and yet others will be basic machines little different than the cars of the late 1990s. It is this complex landscape of old and new technologies which is the reason why future devices must be built to be adaptable, extensible and able to communicate between each other.