They fed a neural network billions of sea-surface elevation measurements taken from 172 buoys located off the coasts of the United States as well as the Pacific islands.
Each of these buoys has been recording for a different period of time, says Breunung, but piecing them all together provides “880 years of data” on ocean waves, including thousands of rogue ones that were substantially higher than waves nearby.
Training their computer system on all this data eventually allowed it to recognise the waves that occurred before a rogue wave happened, and to distinguish them from waves that weren’t followed by this kind of event.
“With this approach, we could predict three out of four rogue waves,” says Breunung. “I was surprised.”
thanks for pulling out the key part. It’s funny, there have been so many articles over the past few weeks about how AI/neural networks have been overblown, but it’s clear that there are some cases where AI can be transformative. This is one really amazing example.
Imo it’s both overblown and very impressive. Deep neural networks are capable of many things that we didn’t even imagine 10 years ago. We’ve made huge leaps.
The problem is that every company is putting “AI” in everything and techbro’s and managers are heavily overvaluing the technology. Most companies don’t need AI. In many cases there are way better methods to do the thing they want to do. The fridge or washing machine doesn’t need AI, the website of whatever company doesn’t need an AI assistant, and most people don’t need an AI accelerator in their laptop or phone.
