Your brain runs on roughly the power of a dim light bulb, yet outperforms machines that consume the electricity of entire power plants. That striking mismatch is the launching point for neuroscientist Henry Markram of EPFL Lausanne, who argues that today’s AI is pursuing intelligence “180 degrees opposite to how the brain works”.
Markram explains that today’s approach to machine learning relies on vast amounts of historical data and expensive computing power to identify patterns. This is an indirect way of trying to understand the world, Markram emphasizes, and contrasts with the way humans capture knowledge. “The brain is not an observer AI. It’s a participant AI.”
From this premise, he introduces “causal AI” as the next big step in artificial intelligence: systems that learn dynamically, require far less data, and improve continuously without costly retraining.
Through the Open Brain Institute, aims to make this research widely available and drive progress. “We’ve built the full stack for the causal AI paradigm, where you can now harvest $100 billion of research research on neuroscience, continuously through the Open Brain Institute”, Markram says, who also co-founded the science platform Frontiers.
“All the code is open”, he emphasizes. “All the data is open. Everybody can be participating in all of this, understanding the brain, building the brain, exploiting the brain.”
Watch the video to explore this talk in detail.
