Our greatest risk is not AI but its opposite: Human stupidity. Our greatest opportunity is not AI but its complement: Human values

“Human stupidity” is Artificial Intelligence’s greatest challenge and risk. It’s embedded in machine learning’s essence. Artificial Intelligence is not “right”. AI is just super-efficient. AI systems learn from data, and for this reason they are prone to picking up our “human stupidity” and our human biases, which data already presents, embedded in them.

As a practical example, a hiring system built with Artificial Intelligence and trained on existing employment data, in most organizations in our Western world, is today likely to “learn” that being a woman and being non-white correlates poorly with being a senior manager. That’s not a fault of AI, but it is “just” the sad result of today data. Most leaders in most Western organizations are today white men. That’s what AI would find in the data today, and that’s what AI would eventually learn from us.

At the same time, machine learning’s very nature may also be bringing us to think about fairness in new and enriched ways. Artificial Intelligence may give us concepts, a new vocabulary, and new tools that enable us to address questions of biases and fairness more directly and precisely than any time before. In addition, as AI has no subconscious, once we intentionally remove biases from the AI algorithm, for instance gender and skin color in the previous example, AI does it for real, in a consistent way,

Therefore, the greatest risk I see today it is not Artificial Intelligence but its opposite: Human stupidity. The greatest opportunity I see today it is not Artificial Intelligence but its complement: Human values.

I look forward to reading your experiences on this important topic. Thank you sincerely for your time and for any input you may provide. We really need lots of different perspectives to reflect on this.