AI is not a ‘Best Buddy’

The phenomenon of seeing AI as a “best buddy” is particularly widespread among younger people. Seventy-two percent of teenagers in the United States have already used AI chatbots as a substitute for friends.

Some executives also believe that using AI could solve all problems in a business context. One thing is certain: in clearly defined fields of application, algorithms already achieve incredible things today. But when AI applications write software code that is transferred into production systems untested and without expert supervision, greater efficiency can quickly turn into greater risk errors and security vulnerabilities can arise.

That can quickly bring entire IT systems to a standstill. Apparently, many people prefer to ignore this danger for now. Fixing the resulting errors could become a costly affair. So, we urgently need to differentiate:

Which problems should I trust AI with? And for which must I, as a human, remain in the lead?

I am not writing this text to argue against the use of AI. That would not only be foolish but also ridiculous. But the often-promoted unlimited benefits should be put into perspective from time to time. It helps to pause occasionally and consider not only opportunities but also risks.

That is sometimes not so easy. Anyone who regularly uses ChatGPT or its improvements knows its advantages. It cites sources you might never have thought of. It makes connections you might never have considered. You almost feel as if you are exchanging ideas with an all-knowing person.

How enriching.

But that is exactly where the temptation begins:

“We mistake the echo for a soulmate, the algorithm for an intellect.”

As long as you are familiar with a subject, you can usually evaluate the suggested sources reasonably well. But where contextual knowledge is lacking, it is easy to go astray. Prompting must be learned. And many people forget while they are still typing that the algorithm sometimes invents answers simply to “please” the person writing the prompt.

How important context is was recently described to me by a young social worker. He and his team look after children and teenagers after school. In the past, the focus was on lunch, homework, games, and sports. Today, conversations often begin with helping the kids interpret—or correct what they have “learned” through chatbots and social media.

It quickly becomes clear which “bubbles” the kids are moving in:

It is exhausting, but also rewarding, because the young people engage in the debates and even enjoy the discussion rounds with real people.

In everyday work life, it is also clear that complex problems cannot be solved by AI alone. The availability of high-quality data remains a challenge, and human judgment is crucial.

The knowledge and experience of employees, their ability to interpret information, and not least personal interaction all remain relevant. And AI also reveals what kinds of personalities we deal with at work:

Can they reflect or do they only react?

Are they trapped in the vanity of their own echo chambers, or do they retain the ability to think independently?

One thing will probably remain the same even in the age of AI: It is the human being who makes the difference!

For as long as people can remember, they have wanted to know what the future holds. Yet in the age of artificial intelligence, almost all predictions are now failing. Developments are progressing faster, more broadly, and more profoundly than we ever imagined—and everything suggests that this pace will continue to accelerate. AI is increasingly entering our lives and communities.

Precisely for this reason, it is so important that we actively engage with the interface between humans and AI. How can we empower people so that AI supports them rather than replaces them?

Let me add a bad news which appeared today: Happy (and safe) shooting: AI chatbots help plot attacks!!!

From school shootings to synagogue bombings, leading AI chatbbots helped researchers plot violent attacks.

In other words: AI is not a ‘Best Buddy’; we must stay in control.

I am looking forward to your comments; contact me at [email protected]