What happens when we are reduced to an algorithm?
This is the fourth of several ‘Innovation Perspectives‘ articles we will publish this week from multiple authors to get different perspectives on ‘Thinking about the future: what big innovation do you expect within 10 years?’. Here is the next perspective in the series:
by Rocco Tarasi
I think some of the most significant innovations of the next decade will come from the increasing amount of digital information that is available about all of us, resulting in a more perfectly predictive economy – probably benefiting us as consumers, but equally hurting some of us that might be employed as ”producers”.
Consider just a small sample of the digital information available about me, not just in what I have done in the past, but also what I will do in the future.
- Google knows a LOT about me, from the content of my emails, to the articles I subscribe to on Reeder, to where I go from Google Maps, to the websites I search for and look at, to the stocks I own from Google Finance, to what I’ve done and what I will be doing in the future from Google Calendar.
- Amazon knows what I buy, how much I like what I’ve bought from my ratings, and also what I will buy in the future from my Wish List.
- Pandora knows what music I like, Netflix knows what movies I like, and both have algorithms to determine what music and movies I will like in the future.
- Facebook knows a LOT about me too, based on my Likes, my profile, who my friends are, my communications between my friends, and the events that I have attended and will attend in the future.
And that is just the tip of the iceberg in what is available about me online. We are in the very, very early stages of making that information useful, but over time the ongoing correlation of all of this historical and predictive information will result in an increased ability to predict what I will do, where I will go, what I will buy, and who I will like – all probably before even I do. Heck, Facebook could even know my sexual orientation before I do.
Now, the most obvious application of this information abundance is more targeted advertising, and that is of course happening today. But the real innovation will come further up the production cycle, before companies have even made their products. Information abundance provides the fuel for more and more accurate predictive analytics. And in that “perfectly predictive” world, companies are only producing products and services that they already know will be hits. No more guesswork.
Some businesses are operating under this “guaranteed” business model today.
- American Idol’s voting system for contestants virtually assures that the winner has a hit album – the audience has effectively said that they will buy it before it is even produced.
- Threadless uses a voting system too, to decide what t-shirts to produce based on user submitted designs – they only produce those with the most votes, and as a result they sell out every shirt they produce.
But these models are decidedly “version 1.0”, requiring explicit user action (voting) to determine an actionable outcome. Version “2.0” predictive business models exploit all available information to accurately predict outcomes without requiring user actions. Stock analysts don’t need to survey you anymore to predict Walmart earnings; they just use satellite photos to count the cars in the parking lot.
We are in an age today where the raw information is being created and the underlying “plumbing” is being laid, as MG Siegler at Techcrunch wrote recently regarding our social connections:
- He [Mark Pincus of Zynga] expressed excitement in the fact that while right now, many of these companies are still working on building out the infrastructure for this social experience, in five years, all this plumbing, as it were, will be in place. Then the real fun can begin… The others on stage [announcing a new social venture fund] echoed this sentiment, suggesting that today is about laying the groundwork for something much greater to come.
Information abundance and predictive analytics have implications for us not just as consumers but more importantly as “producers”. Companies are organized today under the assumption that they operate with imperfect information, and they are staffed based on the resulting assumed inefficiencies. Not all decisions will be right, not all products will be hits.
But what happens when information does become perfect?
If your job or business today is based on imperfect information, then your job will change as information becomes more perfect. As an example, venture capital firms are starting to experiment with technology that makes more accurate predictions of start-up success. Considering that venture investments today succeed maybe 20% of the time (depending on how you define “success”), how many venture capitalists are needed when it is easy to predict who will succeed and who will fail?
What exactly will happen when we are all reduced to an algorithm?
Image from Flickr user Joao Trindade
You can check out all of the ‘Innovation Perspectives‘ articles from the different contributing authors on ‘Thinking about the future: what big innovation do you expect within 10 years?’ by clicking the link in this sentence.
Rocco Tarasi was an accountant, investment banker, and CFO before becoming a technology entrepreneur.