When you purchase through links on our site, we may earn an affiliate commission.Heres how it works.
New ways to learn
Machine learning is old news compared to foundation models.
Who better to ask than Juan Bernabe-Moreno, Director of IBM Research Europe for Ireland and the UK.
We talk about months, even years.
What’s going to be the performance?
That’s something that’s in the making.
But the first results, they look very promising.
According to Bernabe-Moreno, there are three perspectives.
The idea is to use the technology to detect a pattern.
This pattern is constituted by repetition.
So the more extreme weather events we have, the easier it is to find the pattern.
Given the coverage here, if I know that this measure has already been put in place.
Even outside of insurance, the applications are seemingly endless.
For example, for property development.
Which climate adaptation measures are important?
Which ones are in place?
You really get to see how these things happen.
So, for property development it’s very important.
You have another way for agriculture.
How is land use evolving and what are the consequences?
And what’s the right throw in of crop?
And what’s the right irrigation strategy?
So there are many.
That’s the beauty about this foundational model.
It is hoped that the foundation model will be able to automate and scale climate risk assessment processes.
So they can quantify and they can take some measures, for example, understanding the flooding effect there.
That’s very new because you see that end to end.
This is what most excites Bernabe-Moreno.
We are making these efforts quantifiable.
Now you’ve got the option to really track that.
It’s very difficult to say We planted so many trees.
But where are they?
What’s the effect?
How many did we lose, right?
And now, actually we can measure, and that’s for me a game changer.
