The Internet of Interactions
Information Architecture, Open Data,
and the Future of Smart Cities
When we read Internet of Things commentary by our colleagues in the consulting profession, we’re often perplexed by the two-by-two charts they have showing company activity. The upper right-hand corner of the upper right-hand box—the “magic quadrant”—is usually empty, and the favored firms are clustered in the upper right-hand corner of the lower left-hand box, with the clear implication that the IoT is still not ready to “cross the chasm.”
And yet we clearly see multiple parallel technology developments (cloud, mobile, AI, machine learning) increasingly reinforcing and accelerating one another, creating the coming “combinatorial explosion” that we’ve often said will characterize the IoT. With all that evolving capability, how can the IoT still be stalled? And what will be the effect of all that technology on human beings? Will it offer a democratic use of technology, or democracy mediated by high-tech?
It may be both. As we’ve said many times, scalable design concepts will inevitably be imposed top-down by skilled practitioners, but the inputs will always be as funky and bottom-up as humans can dish out—a great virtual machine of billions of interacting nodes. Nothing typifies this Internet of Interactions better than the Smart City, where the most advanced designs and engineering meet the great mess of earthly reality.
As we contemplate the future of our great cities, one thing’s for certain: We can’t wait to achieve consensus on every detail before we begin. Consensus of that sort is something we’ll never have. In the words of Usman Haque, a far-sighted architect, artist, and IoT innovator, “We can collaborate even when we don’t agree on everything.”
But how? We think it’s by learning what the Internet (not the Web) taught us: Deploy simple, smart engineering that can scale into the unknown future—as smart as the bit, the byte, and the packet have been—and always be committed to opening up possibilities for all stakeholders, not shutting possibilities down.
HUMAN REALITY WILL ALWAYS BE A MESS
In the IoT, and especially in smart cities, not all stakeholders will be human beings. In fact, the vast majority won’t even be complex machines. They’ll be simple sensors and actuators. The fundamental premise of the IoT is that everything, right down to those sensors, will be connected and able to talk to everything else. In simple and even compound IoT applications (e.g., smart buildings), this hasn’t posed a great problem since the devices and data are typically owned and controlled by one organization. And the data themselves are in known formats, gathered into one central repository, and operated on with consistent engineering.
But when we venture into true IoT territory like smart cities, the combinatorial explosion quickly surpasses anything we’ve ever dealt with successfully. The reason is that in the real world most data continue to be siloed, stored in their own realms, and this is not going to change in the future. Human reality and the data describing it are a messy affair to say the least, and that reality is complicated further by the need to respect personal and organizational privacy restrictions (e.g., HIPAA).
Clearly, centralizing the world’s data in a single repository is not going to happen.
Usman Haque believes that forward-looking cities are now seeing the benefits of distributed IoT data systems, which leave the data where it’s created and owned. That approach also has the additional benefit of keeping the cities out of multi-decade agreements with Big Data companies whose grandiose plans evaporate every time the marketing wind changes.
DOING THE SAME THING, OVER and OVER
People familiar with our work know that we’ve been critical of the Web’s lack of a real information architecture. This has struck some of our readers as odd, since they think the Web and the Internet are the same thing. But whereas Internet Protocol (IP) is a timeless piece of scalable engineering, the Web is just a great formless blob one layer up. In IP, the basic, irreducible information-unit is the packet, just as the bit and the byte are the basic units of computing itself. What’s the basic information-unit of the Web? The link, the page, the domain, the server?
You can’t say because there isn’t one. If anything is supposed to be foundational on the Web, it’s the idea of hyperlinked text. So try to follow links more than five years old on most Web sites. See the problem? The Web is bad engineering. It wasn’t designed to hold all the world’s interconnected information for all time. In fact, it was originally designed to share scientific papers. Only later did some entrepreneurial types see that they could use Tim Berners-Lee’s HTTP protocol as the basis of a public network.
Now, less than thirty years later, we’ve dug ourselves a hole so deep we’ll never get out of it. In the process, we’ve made the Web do some fantastic things—with kludge after kludge after kludge. We’ve kept replicating our error all the way up, until now we’re at “the cloud.” Did you ever wonder how it got to be called “the cloud” when it’s composed of completely separate, walled-off things? Is Amazon’s cloud connected to Facebook’s cloud, or Apple’s cloud, or Microsoft’s cloud, or Saleforce’s cloud? Not at all. In diagrams from the old days, a picture of a cloud meant “the Internet.” These days it just means “our servers.”
Since it’s walled gardens all the way down, maybe we should just call it “the wall,” as in, “the thing we’re about to hit.”
A TRUE INFORMATION MARKETPLACE
What the market really needs are platforms that anticipate developers’ and users’ toughest challenges—from interoperability and latency to database dependency and user complexity—as a group of problems that can be addressed by a single, unified, scalable software solution. The rigid and fragmented nature of software offerings available today make it extremely difficult to develop effective Smart Systems applications. The required capabilities for next generation platforms will need to be organized around a data and information architecture where there are no artificial barriers between diverse data types, and that facilitates free flowing data discovery, data fusion, and collaborative application development. Acceptance of this reality is essential to the effective design of future-proof platforms.
The refusal to be stopped by reality’s boundaries brings us back to Usman Haque, who has made an extraordinary career by refusing to respect the disciplinary lines separating architecture and art from high-tech and entrepreneurship. Haque has co-founded a number of prescient IoT ventures, including Pachube (a generalized data platform for an open Internet of Things, launched in 2008 and acquired by LogMeIn in 2011), Thingful (a search engine for the Internet of Things), and Umbrellium (technologies to support citizen empowerment and high-impact engagement in cities).
He has his own angle on the failings of the Web, starting with the decision to make all HTTP links unidirectional. The Web allows you to create a “link” to any webpage, but the owner of that page does not necessarily know that you linked to it. Similarly, you have no idea who or what is linking to your webpages. This decision contradicted the design principles of Ted Nelson’s almost unbelievably foresighted Project Xanadu, for which Nelson had invented the terms “hypertext” and “hyperlink”— in 1960, believe it or not. (A friend who saw a private early demonstration of HTTP by Tim Berners-Lee tells us that the Web originally included bi-directional links, too, as well as the ability to modify pages remotely as opposed to just browsing them.)
“This has had an important and profound effect on how we use the Web,” Haque writes, “because the unidirectionality has, arguably, led to everything from search engines (built on the fact that links are hard to find) to massive social networks (built on the value of sharing links and end-user content) to copyright battles, advertising business models and ‘fake newsʼ.” You might even “trace how the unidirectionality of links has led to a ‘sharingʼ economy thatʼs not really about sharing after all.”
It’s typical of Haque’s boldness to invoke Ted Nelson. Xanadu is, after all, the world’s most famous vaporware. Even in a time where we openly seek “unicorns,” Xanadu has been a decades-long wild goose chase. Many organizations and companies, including Autodesk, have tried and failed to bring it into being. In 1988, Autodesk acquired 80% of Xanadu and brought Nelson in as a fellow. John Walker, the founder of Autodesk, said, “The Xanadu design is unique in that it rejects from the outset all limits on generality, capacity, and extensibility. Implementing it in its entirety will be difficult, protracted, and expensive, but no system less ambitious can be as useful, as powerful, or as important for the long term.”
Today, Usman Haque wonders what kind of “Internet of Things” we might arrive at if we followed Nelson’s 17 “rules for Xanadu,” in which we find these futuristic precepts:
» Every Xanadu server is uniquely and securely identified.
» Every Xanadu server can be operated independently or in a network.
» Every user is uniquely and securely identified.
» Every document is uniquely and securely identified.
» Every document can have secure access controls.
» Every document can contain a royalty mechanism at any desired degree of granularity to ensure payment on any portion accessed…
All the way back in 1960, Nelson didn’t just invent the terms “hypertext” and “hyperlink.” He proposed a global, open information marketplace, the very thing that our most visionary tech leaders now see as the solution to the inescapable mess of human reality.
If the Internet of Things is ever going to “cross the chasm,” those numberless privately-owned data-silos will have to be made discoverable, searchable, and actionable in real-time. Unfortunately, we still haven’t approached a solution to the raw engineering challenges of this, never mind the contractual, political, and privacy enigmas. No wonder the Big Data companies stick their heads in the sand whenever this subject comes up, or whine about their need to house and control the data in their own “clouds.”
For Smart Cities and the IoT of the future, the “crazy” ideas of Ted Nelson might be the inspiration we need.