There’s a lot of movement in the tech space today, as developments in AI, machine learning and now deep learning are coming at a pace best described as rapid-fire. There’s a substantial amount of buzz around that last term, though—the newest to the group of powerhouses with the potential to change everything.
Let’s examine what exactly makes deep learning so promising and explore what it means for the enterprise.
What Exactly is Deep Learning, Anyway?
Deep learning falls under the umbrella of artificial neural networks (ANNs), which, essentially, are clusters of virtual neurons created to learn from data sans human supervision. If this sounds a whole lot like what you know of machine learning, that’s because it is—both techniques extract statistics and classify results after looking through large amounts of data. Think of deep learning as an ANN on steroids, working harder and faster because the process is now enabled by huge quantities of data and ever-quickening computational capabilities.
Deep learning facilitates any activity requiring analysis of large amounts of data. Think about that for a second . . . large amounts of data? Where have we heard that before? (Big data, anybody? More on that later.) Now you see why the implications of this technology are staggering. Thus far, deep learning has aided processes including image classification and language translation, and it can be used to solve any pattern recognition problem—remember, this is all without human intervention.
There are a significant number of startups currently aiming to get a slice of the deep learning pie. The breadth of their industries is indicative of the reach of this technology, as areas of expertise range from genomic medicine to social media analysis to agriculture.
Today, many investors are searching this pool of startups for companies with the ability to leverage deep learning into a self-replenishing data network of sorts (See Figure 1 below). Winning companies will rebuff relying on outside APIs and instead invest in leveraging in-house AI capabilities. What this means is these companies can build proprietary data sets that will train the algorithms: More data, more customers, more insights—rinse and repeat. Not only is this self-replenishing, but insights will get deeper as the deep learning algorithms get smarter.
Figure 1. Source: Venturebeat
How Will Deep Learning Change Marketing?
Deep learning can mean mega opportunities for marketers. After all, successful marketing initiatives connect with audiences and compel them to action in some way. Deep learning connects the dots between data points to deliver brands the insights they need to tailor their marketing initiatives to specific cohorts within their target audience base . Here’s how you can expect the magic to happen:
Segment on social. Deep learning can be used to segment audiences across a number of social media platforms quickly and efficiently, giving marketers a wide-lens into what campaigns are working, what campaigns need to be modified, and who’s talking about them. It can also give your brand characteristics of consumers in each of these sub-groups for even more insight.
Recommend what’s relevant. After audience segments are in the books, deep learning can help brands identify what types of content resound with what sub-groups. And while of course it can help with easy things like what resonates best with different audiences on social, that just scratches the surface on what deep learning and the insights it can provide over time will be able to deliver. Think about how this might impact your lead gen initiatives and power your business development efforts, how it might impact your customer service team, your R&D efforts and beyond.
Target ads the smart way. Content recommendation is one thing, but can deep learning actually suggest what ads will perform best (and where)? Why yes, it can—and in real time, no less.
I don’t believe deep learning is going to be a game changer for the enterprise—it’s going to be bigger than that. It is on course to simply become part of the game itself, slowly weaved into our vocabularies and our strategies the same way big data eked on the scene circa-2001 with Doug Laney’s “3Vs.” In the example of big data, what started out as novel became so ingrained into how companies function that it moved from buzzword to business staple. It’s no longer a question of who is using big data—it’s who is using it well? Who is getting the most insight? We’ll soon be asking ourselves those questions of deep learning—who is leveraging the technology to bolster dataset quality, domain expertise and marketing initiatives? Same game, different players.
Where is your company with regard to deep learning? Are you still observing and learning, or have you taken the leap? I’d love to hear your thoughts.