Every decade has one or more (usually more of late) technology trend(s) that the big tech vendors of the time miss. In in the 1980s it was PCs, client server computing and PBXs that did them. It was the Internet in the 1990s, and search, social media, the public cloud and mobile devices did the most damage last decade. This decade we have a number of major trends from consolidated converged data centers (VCE) and applied social manipulation, to robotics and artificial intelligence (AI).
In an earlier article, I looked at these separately. Here I’d like to look at the combined impact of all four of these trends because they do dovetail very nicely.
We start with artificial intelligence because that takes people out of decision trees. These are systems that can look at massive amounts of information, form opinions and then execute based on those opinions. While Watson is the poster child for this effort, there are others like Deep Mind at Google, Sentient Technologies, and Beyond Core in the mix. On top of that there are likely a variety of military-focused efforts that we aren’t allowed to talk about.
Virtualized consolidated data centers provide a flexible platform upon which these AIs can reside so that they can replicate or move to protect themselves at will. This segment is experiencing near exponential performance boosts, so once started, turning one of these things off may be near impossible.
Social manipulation gives these systems the power to manipulate large populations through the control of news and information. In effect, it should allow these huge systems to change human sentiment, making people easier to control. Instead of reacting to what customers want, these system will, with significant increases in analytical accuracy, be able to tell people what to think and how to act.
Finally, robots will become the proxies for the AIs, allowing them to move from their safety bunkers and increasingly to become the primary interface between these systems and the people they were designed to serve.
This comprises a solution that would have broad utility, starting from being able to drive customer needs and fulfil them to creating systems that interact with people in the nearly the same way that people interact with people.
There are two possible paths for such a solution, and they aren’t necessarily mutually exclusive. One is to model it after the old IBM mainframe in that people lease the system and the different components are either hosted or replaced on a regular cadence. The vendor that provides this at scale not only gets a very nice annuity, it increasingly controls by proxy the collected opinions represented by all its customers’ customers.
The other path is to sell the parts and allow firms to build the solution themselves. This provides a higher initial revenue stream and a far smaller annuity. This is how IBM changed their leasing model, but the upfront costs can be more daunting and, more importantly, the solution provider loses a significant amount of potential market control.
I expect we’ll see both models at play, but it looks like Google, which has more of the parts than anyone else at the moment, is working toward the first workable solution.
We are going to undergo a massive amount of change in the second half of this decade, and if AI matures as expected, the result will be increasingly autonomous systems with the unique ability to be mobile and interact in human-like ways.
Some call this the Second Machine Age. I call it a more of a revolution. And this revolution will likely go far better for companies that understand the opportunities and risks of this change early, because the ones that understand it late likely won’t be around to fully understand what they missed.
Put another way, in a successful revolution, you want to be part of the rebels and not the empire. Just saying.
Photo courtesy of Shutterstock.