Applications and support levels of machine forecasts

In an article worth reading in Controller Magazine, Professor Dr. Heimo Losbichler poses the rhetorical question: “Is artificial intelligence a paradigm shift in controlling with regard to planning, budgeting and forecast?”

In doing so, he compares the developments and possibilities with, among other things, the developments we are also experiencing in the field of “autonomous driving”. The decisive factor is the human being. But what can machines already do today?

The discussion so far has shown that

▶ AI systems, or machine predictions, are not yet very widespread and in their infancy, but they are of great importance and have great future potential.

▶ The ideal of accurate predictions will remain an unattainable, even in the age of artificial intelligence, but their use, however, can improve the human predictive ability and automate or support the creation process.

▶ Humans do indeed have cognitive abilities that machines do not (yet) have.

So the question is how machine predictions can best be used. Should they replace or complement human forecasts? A distinction might be made between different support levels of “Assisted Intelligence, Augmented Intelligence, Autonomous Intelligence”, similar to the way it is done in autonomous driving.

At the end of the paper there is a proposal for clustering on “priority applications of machine prediction”.