Whither symbols in the era of advanced neural networks?
This work addresses the philosophical and cognitive science debate on the nature of human thought, proposing a new research agenda, but it is incremental as it builds on existing arguments about neural networks and symbols.
The paper argues that modern neural networks demonstrate abilities previously thought to require symbolic systems, such as combining ideas and learning quickly, challenging the view that human cognition is inherently symbolic, while acknowledging symbols' role in framing abstract problems.
Some of the strongest evidence that human minds should be thought about in terms of symbolic systems has been the way they combine ideas, produce novelty, and learn quickly. We argue that modern neural networks -- and the artificial intelligence systems built upon them -- exhibit similar abilities. This undermines the argument that the cognitive processes and representations used by human minds are symbolic, although the fact that these neural networks are typically trained on data generated by symbolic systems illustrates that such systems play an important role in characterizing the abstract problems that human minds have to solve. This argument leads us to offer a new agenda for research on the symbolic basis of human thought.