Researchers at FAU have used AI to analyze brain activity, uncovering that spontaneous brain signals (local field potentials) help predict future events, even in a resting state. These signals influence how thoughts and emotions are processed, providing insights into cognitive functioning. The findings could enhance diagnostic tools and treatments for neurological disorders. Additionally, the study suggests AI inspired by neuroscience may improve machine learning systems, especially in predictive functions.
https://neurosciencenews.com/ai-thought-prediction-cognition-27641/
AI mimics human brain efficiency for advanced cognitive insights
Researchers at FAU Erlangen-Nürnberg have made significant advancements in understanding human cognition by integrating artificial intelligence (AI) with neuroscience. Led by Dr. Achim Schilling and Dr. Patrick Krauss, the study focuses on adapting AI models to function more like the human brain. Traditional AI systems consume large amounts of energy, as they process data through continuous numerical values. In contrast, the brain uses electrical impulses, or “spikes,” which are more efficient.
The FAU team developed a method to modify AI’s artificial nerve cells, particularly LSTM units, to behave more like biological neurons. Their modified models, tested on image data sets, demonstrated performance comparable to current AI systems but with significantly lower energy consumption. This breakthrough could lead to more efficient AI applications, with plans to expand testing to complex data, such as voice and music.
This research bridges the gap between AI and neuroscience, potentially revolutionizing the design of AI systems to make them more resource-efficient and brain-like in their processing
https://www.fau.eu/2024/09/09/news/ai-uncovers-the-secrets-of-human-cognition/
0 Comments