Member Article

Human decisions and machines

Today we hear more and more about Artificial Intelligence and the increasing role of robots in industry and society. I hear many concerns about the safety of these systems, but I also hear worries about jobs and decision-making.

Algorithms are getting better and better are solving quite complex problems. So, in the future, who will make decisions, human or machine? And if an algorithm is making the decision, what happens to our jobs?

Understanding decisions

Let us first understand that there are different kinds of decisions. Some of these are very suitable for automation or machine intelligence, others will belong to human beings for many years to come.

Business Intelligence (BI) is the common name for the practice of integrating, analyzing and reporting with data to improve business decisions. In BI we often discuss three types of decisions.

Operational Decisions. These decisions are quite simple and small in scope. Each one does not affect much of your business. An operational decision could be …

What do we do if we see this error message?

Which supplier should I use for this order?

Do we give this customer a loan?

The results of operational decisions can be influenced by Tactical Decisions, for example …

What is our target time for dealing with error messages? Who are our preferred suppliers?

How much money do we give in loans this quarter?

Tactical Decisions are also influenced by Strategic Decisions …

We are getting too many errors with this software. Although it will be expensive, is it time to move to a new system?

Do we build components ourselves, or do we purchase from suppliers?

Should we be focussing on business loans or instead investing in domestic mortgages?

You can see that these types of decisions are very closely related. And you can see that they are increasingly more complex.

Operational decisions have a clear structure and they are very repeatable. Each one - in itself - has a low value. These decisions are good candidates for automation and artificial decision-making. Your next application for a loan will probably be decided by a computer. Of course there may be millions of such decisions, and getting the inputs and outputs wrong can be very costly. We saw this with the sub-prime mortgage crisis in the USA which affected the world economy. But each individual decision is small.

Strategic decisions are often vague and complex. They may not even have an answer that is right or wrong. These decisions are still for human beings. No one is using a robot to decide which new market to enter, or which company to acquire.

Artificial Intelligence or …?

Should we worry about these operational decisions being made by automated systems?

It is important to remember that we are in control. We can decide which decisions to make. The automated system can only do what it is told.

In fact, I am not very concerned about Artificial Intelligence. In the short term, I worry about Artificial Stupidity. I mean, that human beings too often leave decisions to machines when we should know better. We all hear stories about drivers crashing into shops, or driving into lakes, because their GPS told them the “Turn Left Now.”

Human and artificial intelligence are astonishingly powerful when working together - if the human uses the machine as a tool. We are the masters of our technology, not its servants.

Creativity and innovation

There is another area where human beings have a great advantage over artificial intelligence. We can be creative and inspiring in ways that machines cannot.

We are capable of making surprising decisions. We are also capable of making what appears to be wrong decision, because it feels right to us. As the marketing genius, Paul Arden, said, “If you always make the right decision, the safe decision, the one most people make, you will be the same as everyone else.”

Creativity requires us to take risks, which can mean making decisions that appear wrong. Try programming a computer to do that!

Automation and the Economy

As you can see, I believe that Artificial Intelligence is exciting and very promising. But I also believe it is limited compared to the human ability to handle vague, difficult, complex decisions.

In the 1950s, Ford opened a highly automated factory in Cleveland. An official from the company gave the american union leader, Walter Reuther, a tour and said “How will you collect union fees from these machines?”

Reuther simply replied, “How will you get the machines to buy cars?”

In the future will our jobs all be taken by robots and artificial decisions? I don’t think so. Instead I think our jobs will be more interesting, more efficient and more creative because we will be working side-by-side with artificial intelligence.

This was posted in Bdaily's Members' News section by Donald Farmer .