Is Amazon killing Creativity and Innovation?


Not too long ago, I got into a really interesting debate with a friend of ours.  The discussion centered around whether or not the collection and use of information about our browsing habits by websites like, Facebook and Google to present us with “recommended” information based our past activity was a good thing.  I thought yes.  He thought no.  Here’s the crux of the argument.

I think that most everyone knows by now that Amazon collects data on what you search for, what you click on and what you buy (amongst other stuff).  They use this information in the background with some really powerful technology to make recommendations to you about things you might be interested in buying.  I, personally, have found that the recommendations are often good ones and are things that usually I am at least interested in knowing about, if not buying.   This process is a remarkable one to think about because it really suggests that human behavior might be more predictable and consistent that what we sometimes think.  Back to the  argument.

The point of disagreement in my discussion with my friend was over whether this technology was good for us or not.  My opinion is that I love it.  It think it’s brilliant that Amazon can use technology to tell me what stuff I might be interested in.  Saves me time shopping and exposes me to stuff without my having to find it.  That suits my low patience and seems really valuable to me.  It’s like having a personal browsing assistant who knows you so well that he can show you only things you’d find cool.  That’s handy.

My friend, on the other hand, argued that this technology will lead to other more ambitious technology that will, over time, ensure that we have exposure to less diversity and, as a result, become less creative or able to generate new ideas.  His argument is that part of how our brain grows and we find new ideas or inspiration is through exposure to the new and different.  His concern is that the algorithims used by Amazon and others basically learn things about how I behave and then use that information to search their data to find others who behave most like me.  Then, they show me what other people like me have bought (or found interesting).  Ultimately, it’s creating a easier path towards sameness at the expense of exposure to things that may stretch or grow our perspective or experience.

This argument definitely takes on a little bit of a conspiracy theory feel, but I think he may be on to something.  What if, in the future, my television only shows me shows that other people like me usually watch?  Or my search engine of choice only shows results that other people like me found to be useful.  What if it becomes harder and harder to find things that are unusual or different?  What if my technology just leads me towards everything that is in my comfort zone?  What impact is that going to have on us?

While I love the convenience of this technology today, I can see how this technology could easily evolve to help us as humans to follow our natural compulsions for comfort, safety and homogeneity.  At what expense?  No diversity, no risk, no discovery=no innovation, no creativity, no inspiration.

Or maybe, we are blowing the who thing way out of proportion.  What do you think?  Is Amazon, Google and Facebook on the path to killing our creativity and innovation?



  1. Jason, I’m with your friend on this one (despite being an uber-early adopter and technologist!). I think the algorithms that herd us into groupthink are actually very dangerous. I’m also a bit of a rebel – I don’t want to eat at a particular place just because 5 of my friends recommend it. I want to carve my own path. But the easier it is to be “nudged” in particular directions, the harder it will be to stay creatively independent in our choices.

  2. I recently came across a series of articles about a company called Epagogix, which has become one of *the* go-to firms in Hollywood. Epagogix has an enormous database of Hollywood data: info on movies, directors, plot elements, narrative structure, actors, box office results and trends, etc. And they have an amazing proven track record of taking a development script and a casting list, running it through their algorithms and predicting exactly how much money the movie will make. Their predictions are usually very accurate, and are frequently freakishly accurate (within $100,000 on a few mega-blockbusters). Hollywood relies on them heavily, getting enormous multivariate reports to analyze their script and casting options, all with the goal of maximizing profits. (“Okay… What if we remove the car chase in Act Two and cast Shia LaBeouf instead of Ryan Reynolds? Okay, what does it look like with or without a Jay-Z soundtrack?”)

    I harbor no delusion that what they do is accurate and correct, but it misses a VERY important element. Their data is historical; it’s based on *what people have already seen*. On movies that have already been produced, released, and watched. As a result, the optimization of their algorithms relies on following a formula that has already been created.

    Had Epagogix been around in 1941, they certainly would have predicted disaster for “Citizen Kane”. And yes, while its easy to say that “Citizen Kane” was a box office failure because of W.R. Hearst, the reality is “Kane” was different from any film seen before – narratively and structurally – and required a certain amount of time to develop a taste for it. Epagogix probably would have been accurate in its predictions for “Kane”. But that IN NO WAY merits not making it!

    Once upon a time, studio bosses who only created films based on the bottom lines were either cheap businessmen or cowards. Most of them would at least reserve a few dollars for the requisite “critical success but commercial failure”. In today’s business world, they’re hung out to dry for thinking way, since we have data that can (almost) empirically warn them of financial disaster.

    And there doesn’t seem to be any correlation between “broadening people’s horizons” and “financial success”. So, with no way to create an algorithm that will broaden horizons, there’s no need to gather data for that purpose. Only the heart, luck, and time can tell us when a “broadening the horizons” work will be a success or failure. Data can tell us when something has worked and when it hasn’t. Using data the way Facebook, Amazon and Epagogix do, they’ll reject anything “risky” and continue to laser-focus our existing desires into something that sells… over and over… and over… and over…

  3. Pingback: Algorithms kill

Leave a comment

Your email address will not be published.