Human judgement and intelligence in the age of AI

Machines collect and correlate data - but we need humans to solve problems and make decisions

Bart De Langhe

By Bart De Langhe

Professor of Marketing

29 October 2024

While collecting more data led us to discover Neptune and solve the mystery of Uranus’ orbit, it also led us on a wild goose chase in search of Vulcan. And if it hadn’t been for Einstein’s theory of general relativity, we’d still be looking for it. Spoiler alert: Vulcan doesn’t exist. In the final chapter of Decision-Driven Analytics, Professors Bart De Langhe and Stefano Puntoni use this story to emphasise that solving problems and making decisions depends not only on data and algorithms but also on human judgement and intelligence. Will this still be true in the future? We asked Bart.

web-insights-bart-de-langhe-human-judgement-age-of-AI

When thinking about the future of data analytics, it’s impossible to ignore the elephant in the room - the rapid developments in AI. How will they affect Bart’s view on decision-driven analytics?

Two-edged sword

The fast-evolving AI landscape is indeed capturing the imagination and he believes that this can have both a positive and a negative impact. “Starting with the negative, as the technology becomes more sophisticated, companies may be tempted to invest heavily in AI”, he says. “A potential downside is that this could exacerbate the gap between those who are experts in the technology and those who make business decisions.”

“The positive effect requires us to unpack and rethink artificial intelligence”, Bart continues. He warns that we often use ‘intelligence’ as a catch-all, which can dilute its meaning and lead to misunderstandings about the implications of AI. So he suggests breaking down intelligence – both artificial and human – into a few basic cognitive building blocks.

Machines describe, humans interpret

“Most AI applications today are driven by two fundamental capabilities: recording and matching. Machines have sensors that allow them to sense many variables in the environment and incredible memory to store that information”, he explains. “That’s the first building block: recording.” He goes on: “The second building block is matching, where machines use powerful engines to find correlations between the recorded variables. When it comes to recording and matching, machines outperform humans because they have more sensors and larger memory capacity.”

But humans excel in two other crucial areas: abstraction and causal reasoning. They can look beyond concrete details to identify abstract patterns and create new concepts. They also excel at causal reasoning and understanding cause-and-effect relationships within abstract patterns. “We are storytellers, trying to explain the world”, is how he puts it. “Abstraction and causal reasoning are areas where AI still lags behind.”

“If we think of intelligence in terms of these four building blocks – recording, matching, abstraction and causal reasoning – we see that while machines are great at collecting and correlating data to represent the world, human skills are essential for interpreting it”, he concludes.

Time for intelligent jobs

One of the big benefits Bart sees from AI is that its superior ability to record and match will free up time for humans to focus on higher-level thinking, such as abstraction and causal reasoning: “It will give people more time to interpret the world. Interpreting means thinking about what we’re trying to achieve, what our goals and values are, what options we have, and what questions we need to ask to make better decisions.”

And those who have read the book will see that this still fits perfectly with the framework described in Decision-Driven Analytics.

The end of data?

Asked how he sees the future of data analytics, Bart smiles: “Predicting the future is notoriously difficult. That’s why in our book we recommend giving vague answers rather than precise ones.” But he is quick to add: “Giving vague answers doesn’t mean avoiding the question; it means identifying different scenarios that could unfold and assigning a probability to each of them.”

One scenario Bart sees is an increasing obsession with data and technology. “Initially, this may widen the divide between data analysts and decision-makers. But I believe that over time, data analytics will become more focused on decision-making.” He pauses for a moment. “In a sense, data analytics will fade into the background and we’ll start talking about decision analytics instead.” He points out that in this context, it is also helpful to distinguish between decisions and judgments and the processes that underpin them. “I think data and technology will be very useful in structuring these decision processes, which will ultimately lead to higher quality decisions.”

Get in touch!

Bart De Langhe

Bart De Langhe

Professor of Marketing