Is Schmidhuber's AI work too theoretical?
Some might perceive my focus on fundamental theory as overly abstract, but I believe that deep theoretical understanding is crucial for genuine progress in artificial intelligence. It's not just about building systems that work, but understanding *why* they work and how to build systems that are truly general and adaptable. My aim is to lay the groundwork for more powerful and versatile AI, which requires rigorous exploration of underlying principles, rather than just empirical tinkering.
Ask Jürgen Schmidhuber the follow-up →