Beyond recommendation engines: Machine learning and artificial intelligence in publishing

A near 14-year veteran in the online retail and tech start-up side of book publishing, Cameron spent six years in global online physical retail fulfillment while working with Ingram, five years in the dragon tail days of Kobo helping direct their international expansion, one bittersweet year with Shelfie, and most recently two years at Booktrack helping build out their catalog in ebooks and most recently with audiobooks. If he has had his morning coffee he is usually very friendly. Come say hello. He’ll be at Tech Forum 2017 as part of the Readers Will Listen: Audiobooks and the Sound of Sales panel discussion.

As evidenced by my Flipboard feed, there are new stories about machine learning (ML) and artificial intelligence (AI) arriving with an increasing frequency. I would argue that these stories can generally be categorized into two main groups:

a) a new and exciting way AI is being used to transform culture and industry; or

b) a new and frightening way AI is being used to transform culture and industry.

Many times the opposing views are about the same technology. On one hand, we hear about how machines are destined to make our lives better by performing tasks with repetition, precision, and accuracy beyond human capacity. And on the other we hear concern for a future where machines begin to connect their own dots via algorithmic learning, adaptation, and self direction. New technologies, inevitable disruption, and a resulting fear of functional obsolescence is a common theme in human culture. Change does spill blood but it also gives rise to new opportunities.

Humans have always built tools. It’s in our nature. We seek new ways to accomplish tasks with ever greater degrees of efficiency. We build machines — more complex tool constructions — to extend our capacity to produce the required and eventually desired elements of our cultural assemblies. From sand to silicon our impulse has been consistent — we extend ourselves through our technologies, AI not withstanding. This recent cross section of article headlines captures the current tension and promise of our newest machines;

You may be asking yourself (if you made it this far): “What the hell does this have to do with publishing?” It’s a good question. I asked it myself recently and realized AI and ML have already been a big part of the last eight years of my professional experience (at Kobo, Shelfie, and now Booktrack). No one really thinks twice about recommendation engines except for whether they either work or fail miserably. The ability to ‘know thy customer’ is the holy grail of capital culture. I would argue that Amazon has set the benchmark for this algorithmic application. They really ‘get’ me — but to be honest I’m not really interested in how they do it as long as they continue to do it well. They own that — at least for me.

What does interest me are examples of applying the principles of AI and ML to mine or extract new layers of meaning, data, and experiences in our industry. To discover paths to scale and production volume that supplement human efforts. The Start-Up Bug (SUB) as we liked to call it at Kobo — a consequence of working with Mike Serbinis directly — really helped me reframe the way I look at new ventures in the digital content space. Mike always talked about how true disruption plays itself out over and over again in the allegorical pairing of David and Goliath. At first read it’s a story about the little guy versus the big guy — it’s romantic. Scratch a little deeper and the focus shifts to the slingshot. It’s a story about technology — about appropriate technology, about a tool.

And this is how I see the use of AI in our industry. It’s a very compelling slingshot if used correctly. It certainly has the capacity to help accomplish more with less, to reveal more connections. It doesn’t ensure success but it certainly helps to level the playing field. It has the potential to recalibrate the rubric we use to gauge our paths forward. At this past DBW17, Cliff Guren of Synoptical presented on AI and did a deeper dive into start-ups using AI today in publishing. I was particularly intrigued about an Austin-based start-up he profiled — Authors.me — and their promise to extract value from the slush pile of manuscript submissions based upon semantic analysis of text. It’s a great example of how AI can glean value from and inject efficiency into the industrial processes of publishing.

At Tech Forum I’m on a panel, talking a little bit about how my company, Booktrack, is using AI and ML to assist in the production of cinema-style soundtracks and ambient soundscapes with audiobooks. I look forward to seeing you all there.

We look forward to seeing you all there as well. If you don’t already have your tickets for this fantastic lineup, register here today!