There comes a point in most programmers careers where they make a startling realization. Computer programming has nothing to do with mathematics, and everything to do, ultimately, with language. It's a sobering thought. The art of computer programming largely involves the creation of and manipulation of text at the level of the individual character, at the level of the word, the line, the paragraph – and from there to the next level of abstraction:
What is the definition of things, or of thing? What is data and metadata, and meta-metadata, and is there a point where the abstraction ever ends? What are functions, not in the mathematical sense, but in the philosophical one? Can one ever describe a thing completely, or failing that, can one describe it adequately? How will we know when we've reached good enough? What exactly is the meaning of meaning?
I am beginning to grow old as a programmer, and perhaps becoming too readily distracted by such fascinating questions to comfortably do the task at hand. It was this sense that I had walking though the Semantic Technology Conference in San Jose this week, In the vendors booths fresh-faced young men and women were showing off the latest in Semantic products designed to ease the workload of the all too busy ontologist or knowledge engineer, databases designed for rapid access of RDF triples and the latest in Semantic entity encoding, editors for building SPARQL queries and tools for the objectification of OWL constructs.
The conference was surprisingly full, filled with people who do deal with ontologies for a living, who are trying to objectify those very OWL constructs, who can speak fluent SPARQL and the four or five (or perhaps six) dialects of RDF and who deal with sophisticated rules engines and semantic inference engines – and yet they weren't the only ones.
There were a surprising number in the sessions, the halls, the floor - more than 1200 people who, often with the blessings of supervisors (or themselves investors or CEO-types looking to make a deal or pick up a fledgeling computer scientist or two), were seen as often networking in the halls many from the government, tasked with the rather daunting task of turning the entire US Federal Government apparatus into a giant data repository.
There were publishers there, looking for an edge to manage their information flows and gain just that critical bit of an edge against the financial and information deluge that's swamping their industry. Barbara McGlamery, Metadata Architect for Time Inc, made the rather startling statement that she now uses RDF repositories for managing the information needs of her company, because information is too multifaceted, too flexible and too interconnected to be stored any longer in a relational database. We are discovering, it seems, that language, that thing which most of us take for granted, is perhaps far deeper than any of us truly realized.
There was a buzz at the conference, too, a rarity in this time of budget tightening and pink slips. One person put it succinctly – “it feels like it's 1994 all over again, and we're on the edge of incredible possibilities”. The buzz was in the products – Wolfram Alpha, computer “search” by way of symbolic manipulation, Bing, Microsoft's new search offering that relies far more on semantics under the hood than may be apparent on the surface. Siri, a personal assistant capable of anticipation, a personal portable Jeeves. Google releasing the bombshell (at least here) that it was now setting up the world's largest repository of triples.
The buzz was in the sessions, attended far more by serious men and women in suits than by young PhD candidates shepharded by their professors (though the latter were still there). This was the edge, no longer the theoretical province of academics attempting to answer deep-seated questions about computation and philosophy, but the realm increasingly of entrepreneurs staking out an early position in a potentially huge market, of knowledge managers within sprawling organizations sensing that this technology might actually be the one to solve the myriad problems of managing the growing tsunami of information within those organizations.
A reality check needs to be cashed here. Hype and buzz may be all to the good, but the greyer beards in this crowd still flinch at words like “expert systems” and the dreaded phrase “artificial intelligence”, even though, by the standards of even a decade ago surpringly many of the goals that were set out for those yesteryear future tech have actually been realized. As Carla Thompson, Guidewire analyst and moderator of a panel of search luminaries, noted wryly “Have we reached HAL yet? Are we close to 2001?” No one on the panel, not surprisingly, took the bait. HAL casts a long shadow.
Wandering the trade floor, I found myself looking at the drag and drop ontology editors, the triples stores and the entity extraction technology, and the thought intruded that this is an industry that has been inward looking for too long, has tried solving problems for perceived needs in an oftentimes very narrow domain, and all too many suffer from the hammer syndrome – OWL and POWDER and SPARQL are very fancy, very complex hammers being used to hammer very ordinary nails.
That's not to say that the technology isn't important and powerful and exciting – in many ways what is emerging now is a rethinking of the very concept of knowledge. Yet it would behoove those who are seeking to impress the rest of the world with their deep thinking and cleverness to be able to explain what it is that they have done, why it is important, and why it should be adopted by others. The proponents of the Semantic Web need to figure out how to communicate with the rest of the world what they have done, why it's relevant, and what ultimately will be the benefits of adopting the technology.
Yet this week was a good start – a necessary prelude as the technology moves from the stage of invention to the stage of innovation (thanks to Tom Tague of Open Calais for this wonderful observation). The Semantic Technology Conference has become the place for these dialogs, and there is no doubt in my mind that connections made in the hallways and dinner tables this week will be the catalyst for a profound surge in both new Semantic web companies and the rise of several existing ones that have toiled in the wilderness for years.