In August of 2011 the Wall Street Journal published a now semi-famous piece by Marc Andreesen called, “Why Software is Eating The World,” in which Andreesen deconstructed the massive and growing impact of software on just about every aspect of human life.

From retail book sales, to animated movies, to music and so on, software was, by Andreesen’s account, running or “eating” the world and would only continue to affect other aspects of our lives as time went on. This assertion was written in 2011. Ask someone in the taxi business today, and they’ll likely tell you that Marc Andreesen was right.

The macro effects of software eating the world are thought-provoking, but the subtle reasons related to human nature are just as remarkable and are core to the effects of software on the world in 2015 and beyond.

Consider the high level evolution of software. Before the PC, a successful software program would sell for up to one thousand units at a cost of between $50-$60k per unit. With the advent of the PC, one thousand units grew to tens and eventually hundreds of thousands at prices of $50-$1000 per unit. Then came the mid 90’s and the birth of the Internet, which dramatically accelerated software consumption and brought about new business models to support software that a consumer would eventually get for free.

In all of this the primary problem that was chipped away at with one iterative solution and business model after another, was friction; friction on both the distribution and consumption end of the equation. This mitigation of friction started with the PC making it physically possible for the average consumer to run a software program on a personal machine and then took a quantum leap forward with the Internet, which allowed all of us to simply download software onto those same personal machines.

This quantum leap in software distribution and consumption took place not because the Internet connected all of us over a virtual network, although that was certainly part of it, but because it did away with the need to go to a store and buy software, did away with the need to load a piece of software onto a machine via floppy disc or CD, and most importantly changed our perception of software from something opaque and complex to something that was easy to get and use, and seemingly limitless in the value it could provide for entertainment, utility and accessibility.

Fast forward to present day. The ability to stream software, where a program can run on a powerful server and be delivered over the air, coupled with the ever-falling cost of core computing infrastructure like storage and bandwidth, has us poised for yet another quantum leap.

We’ve seen the benefits of streaming audio and video files for sometime. However, it wasn’t until Adobe made one of its most important and game changing strategic bets with Adobe Creative Cloud, that we got a sense of what streaming software applications could mean. Adobe’s Creative Cloud allowed creative professionals not only an easier means of software consumption, but also an environment where making, sharing and delivering creative work became infinitely easier and centralized.

Creative Cloud was a 2012 release and has quickly turned into one of Adobe’s primary growth engines, this in addition to Adobe Marketing Cloud. Between this break out financial performance, tied to the Cloud, by one of the largest software companies in the world and the number of cloud based software companies in general, some would think that we’re already seeing the benefits that streaming software can deliver, at scale. I’d say we haven’t even scratched the surface.

Streaming software, as with the Internet, has once again changed the perception of what software is and does. Something once viewed as an individual utility meant to be run by one person on one device, a software program can now be a collective entity, utilized and shared by many, simultaneously.

Because of this, software applications will eventually cease to be something we install onto a device, and shift to a service we simply call upon when needed. All of our interactions with a given software program will be intelligently analyzed and the software program, and our experience with it, will continue to gain efficiency and efficacy.

Consider this real life example. I’m the CEO of Andy OS (the company web site you’re reading this on) and we use Slack for the same purpose that many companies do. Over the summer we had an intern who had his own Slack credentials so he could contribute to group tasks. Less than one month after he returned to school, and without any action on our part, Slack identified that the intern was no longer utilizing and/or contributing to the collective, and subsequently canceled his credentials and refunded his portion of the cost back to us.

Like a human, the program identified behavior patterns, saw a marked change in those patterns, allowed enough time to verify the change and took the logical actions necessary to keep our experience with the program as efficient and effective as possible. Now imagine that happening with ALL of your software. It will.