Sign up

Will self-managing companies ever become a reality?

March 23, 2018

Machine management vs managed by machines

There is much talk of a singularity event and artificial intelligence taking over the world, but will a computer ever replace Bill Gates?   

As a technology company that provides “Sat-Nav-like” tools for managers, and with Sat Navs being the first step towards self-driving cars, we often get asked the “Skynet” question – will scientists ever enable companies to be run by robots?

The answer to that is a definite ‘no’.

Singularity event not on the horizon

Sure, artificial intelligence has come on in leaps and bounds in the last 30 years, but we are not even close to a singularity event where machines can replicate or surpass human reasoning.

What you have to bear in mind is that algorithms function using structured data, yet IBM estimates that 80% of all data is unstructured and approximately 90% of the world’s existing data was created in the last 2 years.

That’s an awful lot of data that machines are missing out on.

You could argue that I’m being unfair pointing this out because algorithms work using certain parameters and, within these confines, they do a brilliant job. But that is exactly the point when it comes to the artificial intelligence vs. human debate – machines do what they are programmed to do while people can be more creative.

The debate being driven by Uber, Google and Tesla

Currently, the debate on artificial intelligence is being shaped by self-driving cars. As Uber, Google and Tesla race to get the first self-driving car on the market, people assume this is just the thin end of the wedge and will very quickly lead to machines running a company, or even the planet.

Here’s the thing.

Managing a company requires complex skills to navigate ambiguity and interpret human sentiment – it’s often ‘messy’ and unquantifiable. Self-driving cars only need to read physical things – road conditions, other cars, pedestrians – using measurable data like location and speed. Even then, there could be ‘black swan events’ that programmers hadn’t taken into account.

Cognitive dissonance and flying sheep

In 2012, in Melbourne, Australia, a truck carrying sheep crashed on an overpass and this resulted in hundreds of sheep raining down on the freeway below. While a human would overcome the cognitive dissonance of flying sheep, you could just picture a confused Uber robot uttering that old sci-fi chestnut, “Does not compute!”

That is an extreme example of an unforeseen event, but it does help to illustrate why machines won’t be taking over companies anytime soon.

The rise of robotic enterprise

Of course, just like a lot of manual labour jobs were replaced by machinery during the industrial revolution, there will be many administrative functions that can be transferred to machines. However, the more creative and high-level thinking roles will remain in human hands.

This AI/human interplay has been the subject of much research at carmaker Volkswagen in recent years.

Chief Information Officer Martin Hoffman explained how the company had adapted the checklist for autonomous cars to the field of “robotic enterprise”  – using auto-adaptive algorithms in corporate functions and processes.

VW has identified 5 levels of AI control from Level 1 (Manual), where humans make all the business decisions, through to Level 5 (Fully automated), where there is no human input at all.

At Evo we occupy the ‘sweet spot’ of Level 2 (Assisted) on the VW scale – the ‘machine’ offers a recommendation but a human makes the final decision. We do not see our tools as a way to replace human judgment but, on the contrary, as a way to augment it, much as our “a woman’s touch in fashion” experiment has shown…

Further reading:

About the author

Martin Luxton is a writer and content strategist who specializes in explaining how technology affects business and everyday life.

Big Data and Predictive Analytics are here to stay and we have only just begun tapping into their enormous potential.

Hey! Was this page helpful?

We’re always looking to make our docs better, please let us know if you have any
suggestions or advice about what’s working and what’s not!

Send Feedback