At the Brains on Silicon conference in Dresden, journalist Gregor Schmalzried warned that increasing trust in artificial intelligence can erode people’s critical thinking. Schmalzried, co‑founder and host of Der KI-Podcast, cited a Microsoft study showing that critical thinking declines the more people rely on AI and said his presentation focused on how to work with AI agents without losing one’s own abilities.
Threat to Critical Thinking
Schmalzried argued the danger is not only that AI may surpass human capabilities, but that humans themselves may fall behind by relying too heavily on AI. He pointed to a concrete example: the consulting firm Deloitte delivered a 250-page report to the Australian government that included sources later found to have been fabricated by AI.
Correct Answer, Fabricated Reasoning
To illustrate why blind trust is risky, Schmalzried described his own experiment with a chatbot. He sent the bot a selfie taken in a subway underpass and asked it to guess his location. The AI identified the location exactly, but when asked how it knew, it offered a detailed explanation involving steel panels and the color of the asphalt. Only after further questioning did the chatbot admit it had simply extracted coordinates from the image data. The answer was correct, but the reasoning was made up.
Schmalzried used these examples to underline his central point: working with AI requires vigilance to avoid losing one’s critical faculties and to ensure that correct outputs are backed by valid explanations.
