Theory Thursday: Multisperspectivity: The Importance of Holding Contradictions

Art by @basilonmypizza: https://lnkd.in/eF8FkWzN - https://basilhefti.ch/

Is today’s AI imperfect? Yes. Should we ignore it? Absolutely not. The question “What will AI really do for me and to me?” comes up frequently. And nobody has the answer.

To orient ourselves, we look for markers.

We see that current AI still struggles with basic math and conclude: if it can’t do this, then it surely can’t do that. We also see AI highlighting diagnoses doctors have overlooked and think: if it can save lives, then it can do everything. Both conclusions are off.

This is where multisperspectivity comes in: opposing statements can be true at the same time. It happens everywhere, not just with AI. And it bothers us, because many people have trouble holding two contradicting ideas in their heads.

But when we talk about AI’s opportunities and impacts, this ability becomes essential. Kolbjørnsrud’s Relevance Principle captures it well: match the type of intelligence (human or machine) to the nature of the problem.

Lean too far to one side and you miss the point. Lean too far to the other and you miss it again. Progress sits in the balance.

What example of multisperspectivity have you seen recently?

• Kolbjørnsrud, V. (2024). Designing the Intelligent Organization: Six Principles for Human-AI Collaboration. California Management Review, 66(2), 44-64

• Art: https://lnkd.in/eF8FkWzN https://basilhefti.ch/

• I talk about this in my book, https://lnkd.in/eGkQN8t8