My impression is that it’s interesting because it’s good at some functions that deep learning is bad at (although unfortunately the paper doesn’t make any toe-to-toe comparisons), but certainly there’s a lot of things in which transformers would beat it. In particular I would be very surprised if it could reproduce GPT3 or DALL-E. So, if this leads to a major breakthrough it would probably be through merging it with deep learning somehow.
My impression is that it’s interesting because it’s good at some functions that deep learning is bad at (although unfortunately the paper doesn’t make any toe-to-toe comparisons), but certainly there’s a lot of things in which transformers would beat it. In particular I would be very surprised if it could reproduce GPT3 or DALL-E. So, if this leads to a major breakthrough it would probably be through merging it with deep learning somehow.