Who Has the Final Say? The Human Touch in an AI World

How bad would it be if this article were written by AI? Or if the quotes didn’t come from the people on stage, but from an algorithm that analyzed dozens of interviews? Does that change if it’s not an article, but a diagnosis of your symptoms?

There’s a familiar atmosphere on stage. Dina has known Martha and Katherina since the dawn of the internet. It’s precisely during this period that she often thinks back to those days. “Actually, AI reminds me of the advent of the internet,” she says. “Back then, we were mainly focused on the technology. Now the same thing threatens to happen again. Yet what technology does to people and society is at least as interesting.”

Martha, director of the ANP, recognizes that comparison. Precisely because AI is getting better and better at analyzing large amounts of information, there are opportunities, but her profession is also becoming more complicated. “Journalists at the ANP must continue to focus above all on the factual accuracy and reliability of news and information sources,” she says. “No matter how difficult that becomes, this will continue to make all the difference.”

Trust appears to be a recurring theme. Take journalism, for example. Chair Jeroen Smit cites Bellingcat’s MH17 investigation as an example of the power of data. Large amounts of publicly available information made it possible to reconstruct events that had been the subject of debate for years.

“It’s fantastic to be able to determine what happened based on data from many different sources,” says Martha. “But in a growing sea of indefinable information, it’s sometimes hard to decide where to focus your energy.”

That is precisely why the final say remains with humans, she emphasizes. It’s no coincidence that the ANP follows the four-eyes principle and always has a “human in the loop.” “We won’t get very far relying solely on AI,” she says. “And fortunately, people currently place greater value on journalism created by humans.”

A similar issue arises in healthcare. AI is getting better and better at recognizing patterns in scans and medical data—sometimes even better than humans. An algorithm can analyze thousands of X-rays in a fraction of the time it takes a radiologist, but it’s still the radiologist who makes the final call.

When the audience—more than a hundred northern executives and entrepreneurs—was asked whether they would prefer a diagnosis from an algorithm over one from a doctor, about 40 percent chose AI. Sixty percent stuck with the doctor.

According to Katherina, this doesn’t have to be a contradiction. “Many diagnoses are already supported by AI. It helps us recognize patterns that a human might miss.” At the same time, she sees opportunities outside the hospital. “Technology can lower the barrier to accessing care, for example for people with mental health issues, who may then feel more comfortable asking for help.”

This naturally shifts the conversation from technology to entrepreneurship. After all, the technology is there; the revenue model often isn’t yet.

“Entrepreneurs really do want to get started with this,” says Dina. “But it’s not yet easy to develop a profitable business model around it. Especially when it comes to prevention, the system isn’t set up for that yet.”

According to Katherina, that is precisely the point where a choice lies. “We can wait for the government and health insurers to take action, or we can come together ourselves to propose new business models. We’re at a tipping point: something has to happen.”

The audience expresses agreement, with one caveat: a financial incentive is needed, such as a model that rewards results rather than actions. There are also calls for more international cooperation and knowledge sharing.

Dina has noticed another striking development in the use of AI. “Recently, I heard a group of people referring to AI as ‘he.’ Let’s be mindful in our language use that AI is not a human being.” According to her, this is more than just a linguistic issue. The more human-like AI becomes, the more important it is to remain critical of the choices we make ourselves.

That’s why Jeroen argues against focusing solely on the short term. It’s a sentiment Katherina immediately recognizes. “There’s a name for that: Amara’s Law.” According to that theory, new technologies are often overestimated in the short term and underestimated in the long term.