This is because there are small pockets of technologists who are letting their imaginations lead the way. In a suddenly cliché way of saying it, they’re aiming for 10x improvement rather than 10 percent improvement. They can do that because they now have a base set of analytic technologies and techniques that are well positioned to solve, with relatively little effort, whatever data problems are thrown their way.
Here are some themes from our just-concluded Structure: Data conference that I think highlight the promise of data, but also the challenges that lie ahead.
Man and machine unite
Machine learning is already infiltrating nearly every aspect of our digital lives, but its ultimate promise will only be realized when it becomes more human. That doesn’t necessarily mean making machines think like human brains (although, granted, that’s a vision currently driving billions of research dollars), but just letting people better interact with the systems and models
trying to discover the hidden patterns in everything around us.
Whatever shape it takes, the results will be revolutionary. We’ll treat diseases once thought untreatable, tackle difficult socio-economic and cultural issues, and learn to experience the world around in entirely new ways. Maybe that consumer-experience scourge known as advertising might actually become helpful rather than annoying.
That would really be something.
Data science, or data intelligence?
I’m not sure there needs to be a distinction between data science and data intelligence, but the latter does connote a grander goal. It’s about trying to solve meaningful problems rather than just serving ads; about trying to understand why things happen just as well as when they’ll happen. This means learning to work with smaller, messier data than we might like — certainly smaller and messier than the data sets underneath most of the massive web-company data science undertakings.
But just think about being able to go beyond predictive models and into a world of preventative — or even professorial — models. If you know what I like, where I go and who my friends are, it might be fairly easy to predict what I want to buy. Figuring out how my decision to buy something might affect my overall well-being and then telling me why? That’s a little more difficult and a lot more beneficial.
Telling stories with data
Have you ever looked at a chart and wondered what the heck it was supposed to be telling you? Or downloaded a report of your Facebook activity only to ask yourself if all the disparate data points come together to paint a bigger picture? Or tried — and failed — to stop a terrorist before his movement to recruit an army of followers gained critical mass?