Affordances & AI: Human Brain’s Unique Ability
- Researchers at the University of Amsterdam have pinpointed specific brain activations that explain how humans instinctively understand movement possibilities within an habitat.
- The research, led by computational neuroscientist Iris Groen and PhD student Clemens Bartnik, explored how the brain determines potential actions - walking, cycling, or swimming - upon viewing...
- Groen said AI models have "a lot to learn from the efficient human brain" in this area.
Humans understand their surroundings in ways AI struggles to match. A new study uncovers unique brain activations linked to how we perceive potential actions, like walking or swimming, highlighting a critical difference between human cognition and artificial intelligence. Researchers compared human responses with AI models, including ChatGPT, revealing that AI falls short in predicting possible actions within a scene. The findings, from the University of Amsterdam, suggest that understanding brain function, the primary_keyword, could revolutionize AI progress, leading to more efficient and intuitive systems. This research underscores the critical role of secondary_keyword in how we interact with the world, the News Directory 3 team is following the developments closely. Discover what’s next in bridging the gap between human perception and artificial intelligence.
AI Lags in Understanding Environmental Role, Brain Study Shows
Researchers at the University of Amsterdam have pinpointed specific brain activations that explain how humans instinctively understand movement possibilities within an habitat. The study highlights a key area where artificial intelligence trails behind human cognition.
The research, led by computational neuroscientist Iris Groen and PhD student Clemens Bartnik, explored how the brain determines potential actions – walking, cycling, or swimming – upon viewing a scene. The team compared human capabilities with AI models, including ChatGPT.
Groen said AI models have “a lot to learn from the efficient human brain” in this area.
using MRI scanners, researchers monitored brain activity as participants viewed images of various environments and indicated potential actions. The study revealed that specific areas in the visual cortex activate in ways not solely explained by visible objects.
Groen explained that these brain areas represent not only what is visible but also “what you can do with it,” even without explicit instructions. She added, “These action possibilities are therefore processed automatically.”
the research demonstrates that “affordances,” or opportunities for action, are a measurable property of the brain.
When AI algorithms were tested on the same task, they underperformed in predicting possible actions. Groen noted that even when specifically trained for action recognition, AI models’ internal calculations did not align with human brain patterns.
Groen said that even the best AI models don’t give the same answers as humans,showing that “our way of seeing is deeply intertwined with how we interact with the world.”
Groen suggests that incorporating knowledge of human brain function could make AI “smarter, more economical and more human-pleasant,” particularly in sectors like healthcare and robotics where understanding actionable environments is crucial.
What’s next
Future research may focus on integrating these findings into AI progress to create more reliable and efficient systems that better understand and interact with the physical world.