Products in a supermarket are not randomly selected and shelved. Beer and chips are purposefully positioned side-by-side. Sweets are placed in the eye line of kids. High priced brands pay for having their products placed in easy to reach areas. Each inch on a shelve is carefully calculated to be filled with those products that bring the most profit and margin.

Internet companies like Google went even farther.  The position of a link, the color of a button, the right word is selected by A/B tests. Is variation A bringing more clicks or variation B? Do visitors buy more with the first or second word?

That’s not done manually anymore, but done by algorithms. How far that can go is exemplified with ‘autonomous corporations‘, where nothing is left to chance or the humans.

Autonomous corporations replace humans with artificial intelligence. Not the supermarket clerk makes a decision what to put on the shelves, the neuronal network decides what products in what colors on what day on what inch of the shelve at which position is put in store.

“Automation went after blue collar jobs. Artificial Intelligence goes after white collar jobs.” – Ben Levy

While machines replaced human muscles in the past, nowadays computers replace brain muscles. IBM Watson won in 1997 against the chess world champion Gary Kasparov, and 2016 Google’s AlphaGo beat the ruling Go-champion. Fighter pilots are replaced by drones and computer pilots. Drivers by computer algorithms in self-driving cars. Event doctors face competition from AI-systems such as IBM Watson.

At an event this a week by BootstrapLabs, a venture capital firm focused on Applied AI, the topic was on autonomous corporations. Uber’s head of machine learning, Danny Lange, talked about his experiences and trends. While the chess champion lost because of brute force used by IBM Watson, i.e. by pure computing power with predefined algorithms and moves, AlphaGo represented a milestone in the research on AI. No human could pre-program all moves in Go, because of the complexity of the game. Instead the game learned the game over months of play. First by playing against a human, then by playing against itself.

Self-driving cars as well cannot have all the driving scenarios being programmed in by a programmer, they have to learn. Danny showed how Carnegie Mellon’s snake robots learn to move on different terrain and climb over obstacles.

If you take that further, the same approaches can be applied to corporations. An AI-system for a supermarket can test what goes on the shelves. What articles on what height combined with which other articles and in which colors at what time of the day or season generates the most revenue or profit margin?

Uber Eats, a service offered by Uber for its clients, builds on data that other Uber offerings collected. What riders, drivers, and trips were taken, how does that combine with maps? All that together and superimposing restaurant and grocery information on that results in Uber Eats.

Autonomous corporation startups such as Vidora, Roger.ai or Qurious solve parts of the problems that companies have today and make them autonomous in that process. Vidora analyzes what customers will likely churn, or which customer can be most likely upsold. Roger.ai takes care of automatic bill payment for clients, which makes it easier and more predictable on the payments received side for corporations. Qurious looks at how sales reps differ in their approaches and what makes them top or low-performing. The AI-system analyzes their behaviors and sales approaches and supports low-performing in real-time during sales negotiations.

Highly qualified employees who’ve been spending time with routine tasks can now be taken over by AI-systems and executed better than by the human. While automation went after jobs of blue collar workers, now AI is doing the same with white collar jobs. How this will play out in the future and if we are going to be led by entities without humans is to be seen.