Sexy? Maybe. But can they be trusted? Data scientists are the rock stars of the big data frontier. The Harvard Business Review crowned the job the "sexiest profession of the 21st century." During election season, political junkies waited with bated breath for Nate Silver's next blog post. Colleges and universities have tripped over themselves creating degree programs to tap the trend. But, blinded by the ballyhoo, we just might be falling for data scientists a little too hard for our own good.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Pete Warden seems to think so, and he should know. Warden is himself a (sigh, flutter) data scientist. In this week's lead Searchlight item Warden, also a chief technology officer and avid blogger, provides a thought-provoking question about how we use big data, providing an example of his own work to show how it can be misconstrued.
It's a mea culpa of sorts. Warden doesn't cop to doing anything wrong, per se, or suggest malfeasance among his peers. What he does admit is that data scientists aren't necessarily always doing things as right as they could be -- and data Luddites are like us are letting them get away with it. Read the piece to see precisely what you might detrimentally be taking for granted. Bottom line: Everyone is so gaga over this thing called big data, we tend to accept what data scientists tell as unquestionable truths because, well, it's backed up by data.
A more pessimistic sort of person would say the horse has left the barn here -- the potential prospects of mining big data are far too juicy for companies to pull back on the reins. In this digitally driven, customer-focused age, everyone is scrambling for the competitive advantage locked inside their data. But Warden doesn't think it's too late to move in a more responsible direction. As in traditional scientific fields, there ought to be some type of peer review process, he argues. He calls on "real scientists" and social scientists to get on board and work with big data, for companies to be sure the stories their data is telling are true and for all of us to hold data scientists to a higher standard.
Check out SearchCIO's own coverage of
Data scientists say predictive analytics is hard to do
The essentials for building a big data strategy
CIOs should be leading the big data revolution
Data visualization: Should seeing equal believing?
Not surprisingly, given our infatuation with all things big data, the piece touched off an exchange of criticism and feedback on his blog and in the Twittersphere, lively enough for Warden to collect and published much of it in Storify. (Ah, the big data echo chamber.) It even spawned an equally thought-provoking follow-up in The Guardian about the risky business of data visualization, also included in this week's roundup.
Beyond the big data, this week's Searchlight also touches on Google's latest gadget, another reason why we may no longer need wallets, a "very good" app and more.
- Forget about bringing sexy back -- data scientists need to bring accountability back, said one of their own.
- Never say never -- unless you're trying to drive home a point. The Guardian picks up where Pete Warden left off, with a look at why why you should never trust data visualization.
- The concept of bring your own cloud isn't so new, but these reasons to fear it are. Sweet dreams!
- Oh Google, you're making me love you, and I don't want to love you because you're so invasive and creepy in your doings. Alas, your new Chromecast is far too innovative and cool (and cheap!) not to consider inviting it into my living room. (You already know the address.)
- When we start paying for things with our faces, as Finnish start up Uniqul will soon make possible, I fear identity theft could get really ugly.
- And mobile users saw the app and behold, it was very good. Those who want to know how technology can change behavior through consumer psychology and data analytics may want to pick up the Bible (app).