Prospective Learning: Principled Extrapolation to the Future
This work addresses the challenge of learning in non-stationary environments for AI systems and natural intelligence, proposing a foundational shift in learning paradigms.
The paper tackles the problem of learning under dynamic, partially predictable future distributions, reformulating machine learning to focus on prospective learning where distributions evolve over time. It argues that prospective learning is more difficult than retrospective learning and is essential for addressing real-world AI challenges and understanding natural intelligence.
Learning is a process which can update decision rules, based on past experience, such that future performance improves. Traditionally, machine learning is often evaluated under the assumption that the future will be identical to the past in distribution or change adversarially. But these assumptions can be either too optimistic or pessimistic for many problems in the real world. Real world scenarios evolve over multiple spatiotemporal scales with partially predictable dynamics. Here we reformulate the learning problem to one that centers around this idea of dynamic futures that are partially learnable. We conjecture that certain sequences of tasks are not retrospectively learnable (in which the data distribution is fixed), but are prospectively learnable (in which distributions may be dynamic), suggesting that prospective learning is more difficult in kind than retrospective learning. We argue that prospective learning more accurately characterizes many real world problems that (1) currently stymie existing artificial intelligence solutions and/or (2) lack adequate explanations for how natural intelligences solve them. Thus, studying prospective learning will lead to deeper insights and solutions to currently vexing challenges in both natural and artificial intelligences.