AI agents can now decide how far ahead to plan on their own

A new AI learning method lets computer programs decide how far ahead to plan their actions. This means they can react to new information without a human having to tune their planning strategy for each task.

Training AI agents for complex tasks often means a human expert has to fine-tune how the AI plans its actions. This new method removes that step, making it easier to deploy AI in situations where the environment changes often. It could make it simpler to build robots that adapt to new situations on the fly, without needing constant human reprogramming.