The past eight years have been great for the Valley. Before 2008, the valley built technology for large corporations that in turn would use the technology to optimize their businesses. Now the valley is creating new businesses that happen to use technology.
In short order, we overturned the TAXI industry, created the first new car company of note, transformed how we interact with each other, radically transformed how content gets created and delivered, transformed food delivery, are disrupting pay-day loans, and the list goes on.
At the heart of the business models is an understanding of how people interacting with intelligent machines can efficiently deliver services that in the past were too costly to provide.
We have gone from being the disruptors to becoming mainstream.
When Mark Pincus used analytics to help create Zynga, the gaming industry puked all over us. Now, every single game company uses some amount of data analytics to optimize their games.
And that got me thinking.
We have created a bland uniformity in our corporate structure. Our companies look the same, have the same kind of people in it, are structured the same and are leveraging the same kind of technology.
Our venture capitalists are pursuing the same sort of risk mitigation strategies. Distributing their bets across as many good deals as they can find. And yet, the underlying technology structure of most of those bets is similar.
The last time this kind of thing happened was in the banking crisis of 2008 when every single banking company was pursuing the same business strategy leveraging the same algorithms to reduce risk and as a result exposing themselves to the same underlying catastrophic risk.
And startups are doubling down on the intelligent machine model. For example, Zappos is trying to fix human interaction. The Zappos solution is to seek to replace the ambiguity of human relationships with the structure of software systems.
One of my favorite thinkers is Nassim Taleb. His books are difficult to read. And yet he makes a profound point. The more you try and avoid risk, the more robust you make a system, the more fragile it becomes because any remaining weaknesses will obliterate everything.
In our case, the valley is trying to de-risk human decision making using intelligent machines.
There is too much sameness, too much of the same kind of operating model.
And when you see this amount of similarity, you know that this entire world will get disrupted somehow.
My belief is that the limits of intelligent machines are poorly understood. And the faith in the power of those tools will lead to massive amounts of correlated failure. The failures will occur simultaneously because of the sameness. And the effect will be a broad-based failure.
The companies that do disrupt the current valley will be those that understand the limits of machine learning and figure out how to use the human brain, not to make the algorithm more efficient, but to enable the human brain to do things it could not.
What that thing is, is unknown and the timing of the disruption is also unknown. The only thing I am certain of is that both will happen.