Gifs, gifs, everywhere. Once the sole reserve of back-alley Internet forums and Geocity websites, gifs are enjoying a new legitimacy at the helm of the New York Times, the Atlantic, and most recently, Barack Obama's email campaign. I want to argue that this represents not just the evolution of the gif, but the evolution of language itself.

It almost seems like we should revert to the old model of artists and patrons. Lorenzo de' Medici wasn't looking for a return on his investment in Michelangelo. He just wanted some great art. The product was enough of a return. Likewise, Michelangelo wasn't looking to build a business or sell things. He just wanted to make some great art. He wanted to build impactful, beautiful, and timeless products.

In the past, consumers have voted with their dollars. Now, they vote with their eyeballs. Pundits call this "the attention economy". By itself, the attention economy is incredibly efficient. Consumers are quite free to fill their attentions with exactly what they desire. They are not free to do this in markets with money as the medium of exchange. In this world, consumer value is often free to consumers themselves.

When you're on mobile, you're either content to be consumed or you're a distribution channel. A river, or a boat that floats on it. The way people use mobile, they use a distribution channel to get to their content. The most harmful thing you can do to your product is try to build a river when you really should be building boats. Read: news organizations and magazines with native mobile apps that no one ever wants to download.

Out of all the relationships humans have had with each other over the millennia, the engineer-to-user relationship is one of the newest, both in concept and in scale. These days, a single engineer has the capacity to have this relationship with hundreds of millions of people. Really, the last person to have had such a massive scale of person-to-person relationships was Jesus.

One thing I learned during my internship this summer is that the best questions are usually the ones that go unasked. I was assigned to develop a mobile app using a stagnated and complex legacy code base. My mentor was the CTO, and he understandably didn't make himself entirely available to me for questions. He had a company to run, and if my question took him more than a few seconds to think about, he would just push me gently in the right direction and leave it at that.