Pigeons and Artificial Intelligence: A Surprising Connection in Learning

Last Updated on March 28, 2025 by Mitch Rezman

In a fascinating study conducted by researchers at the University of Iowa, pigeons have been found to share a fundamental learning process with artificial intelligence (AI).

While it might seem unlikely that these humble birds have anything in common with advanced machine learning algorithms, the findings reveal intriguing parallels in how both approach problem-solving.

The study, led by Edward Wasserman, a professor of experimental psychology, demonstrates that pigeons rely on trial and error to make decisions—much like how AI systems are trained to learn and adapt. This discovery sheds light on the basic mechanisms of learning and highlights how nature and technology can sometimes converge in unexpected ways.

**Pigeons as Problem-Solvers**

Pigeons are often underestimated, but they are remarkably intelligent creatures. In the study, the researchers presented the birds with a series of visual tasks on a touchscreen. The pigeons were asked to identify patterns and categorize objects based on specific rules. Correct responses earned them a reward, while incorrect ones prompted them to try again.

Over time, the pigeons improved their performance, demonstrating an ability to learn through repeated attempts. This process mirrors reinforcement learning, a method frequently used in AI. In reinforcement learning, machines are programmed to make decisions by receiving feedback—positive reinforcement for correct actions and negative reinforcement for errors. Through this iterative process, both pigeons and AI refine their understanding and improve their accuracy.

**Learning Without Logic**

One of the most striking aspects of the study is that pigeons don’t rely on logical reasoning to solve problems. Instead, they use associative learning, which involves recognizing patterns and forming connections based on experience.

Similarly, AI systems don’t inherently “understand” the tasks they are performing; they rely on massive amounts of data to detect patterns and make predictions.

Wasserman explains that this type of learning is surprisingly effective, even though it lacks the complex reasoning capabilities humans typically associate with intelligence. “Pigeons are not using logic or deep thought,” he says. “They’re just trying things out and learning what works. And that’s very similar to how AI operates.”

Implications for AI Development

The parallels between pigeons and AI raise interesting questions about the nature of intelligence and learning. If simple trial-and-error methods can lead to complex problem-solving abilities, what does this mean for the future of AI? The study suggests that understanding the basic principles of learning—whether in animals or machines—could help scientists develop more efficient algorithms.

By studying how pigeons adapt and learn, researchers may uncover new strategies for improving AI systems. For instance, Wasserman notes that pigeons are highly adaptable and can quickly adjust to new challenges. Incorporating similar flexibility into AI could make machines more versatile and capable of handling unexpected situations.

A Window into Human Learning

While the study focuses on pigeons and AI, it also has implications for understanding human learning. Humans often rely on trial and error, especially when tackling unfamiliar problems. By examining how simpler organisms like pigeons navigate their environments, scientists can gain insights into the fundamental processes that underpin all forms of learning.

Wasserman emphasizes that intelligence exists on a spectrum. “We tend to think of humans as being at the top of the intelligence hierarchy,” he says. “But there’s a lot we can learn from other species—and even from machines—about how learning works.”

**A New Perspective on Pigeons**

This research also challenges common stereotypes about pigeons. Often dismissed as “city pests,” these birds are proving to be valuable contributors to scientific understanding. Their ability to learn and adapt underscores their cognitive capabilities and highlights the importance of studying diverse species.

So, the next time you see a pigeon pecking at crumbs on the sidewalk, remember—they might have more in common with cutting-edge AI than you’d expect!

Conclusion

The University of Iowa study reveals a surprising connection between pigeons and artificial intelligence: both rely on trial-and-error learning to solve problems.

This discovery not only deepens our understanding of how animals and machines learn but also offers valuable insights for improving AI systems.

It’s a reminder that intelligence takes many forms—and sometimes, even a pigeon can teach us something new about the world.

Written by Mitch Rezman and the Windy City Parrot content team.