« Sustainability More Important Than Ever | Main | A Building is Only As Good As Its Foundation »

May 21, 2013

How to Make a Hit? There's a Numerical Formula for That

Posted by Simon Towers (View Profile | View All Posts) at 7:11 AM


Sometimes I think, it's only a matter of time before engineers will come up with software that can produce posts like this one. In other words, I'd only have to plug in a few keywords on what I'd like the theme of my InfyTalk post to be. And a computer will turn out a beautifully written, 700-word essay. No other human involvement would be necessary. In fact, here's an article from the New York Times that supports this hunch. It details how one of the last so-called creative bastions of Hollywood - the movie screenplay and those who write it - is an enterprise that is about to change dramatically. That's because statisticians claim to have developed ways to detect winning formulas of movies using data.

Such a development shouldn't come as a surprise. Look at communications, especially through a political campaign. In that realm - once ruled by intangibles like whether a candidate "looked presidential" - data is now king. Pollsters feed the reams of data to campaign statisticians, who in turn develop a candidate's platform and speaking points. In Hollywood, studios often take gambles on films worth hundreds of millions of dollars. The idea is that by supplying themselves with enough data, they can remove a lot of this risk.

Another industry that data has fundamentally transformed is Madison Avenue. Once run by people revered for their innate ability to know how to cajole consumers to buy products, the advertising world is now led by regression analyses and demographic data. Today a television commercial is designed more by number crunching than by someone's gut creative instincts.

Of course, if data crunchers point to a certain formula as being the winning one in Hollywood, might there be a time when essentially just one film appears time and again? Sure. It will contain all of the elements deemed likely to resound positively with movie-goers. I've heard political pundits say that the importance of data has created essentially one kind of candidate in the world of politics - one willing to finesse his image and talking points to whatever direction the polls take. What's fascinating is that voter turnout in many Western countries (where campaigns use data religiously) continues to slide. So maybe knowing a winning formula for a candidate doesn't necessarily inspire people to leave their homes and vote for him. The same might happen in Hollywood. Do people want to spend considerable money on a ticket if they know they're going to see what it essentially the same story line time and again? It's a question the screenwriters hope their studios will ask.

Whatever you think about how to use data to influence certain tasks that were once solely influenced by creative teams, it signifies something much bigger in the march of progress. I began this post with the subject. I speak of a point at which machine learning takes over many of the tasks we currently think can only be performed by humans. Artificial Intelligence is now a reality. And it's getting more advanced and sophisticated every day. Even a few decades ago we would have been amazed at the notion of a search engine suggesting subjects even before we could finish typing a word. But most of us experience this very thing every day.

On Wall Street, financiers once had to scour news sources for insights into the global markets. Then they could make trades based on what they thought certain segments of the population would be wanting (or shunning). For instance, they began a massive sell-off of shares of ocean liner companies after the sinking of the Titanic. Now that same world is run almost entirely by algorithms that can predict with a reasonable amount of accuracy how many shareholders of a particular company's stock will sell and how many will buy in a specified amount of time. And, like in the world of politics, it's become much more difficult for traders to find tomorrow's runaway stock because of the high efficiency of trading algorithms.

The use of data in every corner of society allows for machines to learn and make moves we once thought of as inconceivable. Like suggesting the elements of a successful screenplay - or an InfyTalk post.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.

Search InfyTalk

+1 and Like InfyTalk

Subscribe to InfyTalk feed

InfyTalk VBlogs: Watch Now

Infosys on Twitter