[D] Current academic research trends v.s. next 5 years

Hello everyone!

As the title suggests, I'd like to ask your opinion (specially for those who are active researchers in ML) regarding which currently hot topics you believe will remain relevant for the next 5 years - and which you see as temporary trends fated to fall off research interest.

For example, two obvious areas of research are LLMs (and transformer-based models, in general) and generative image models (such as diffusion-based architectures).

Personally, I think that Physics Informed networks (commonly …

academic academic research current hello hot machinelearning next opinion research researchers topics trends will