Data sharing and open interfaces for AI optimization. With Paul Patras at Net AI and the University of Edinburgh

Hello everybody and welcome to Sparing Partners.

Today we're again talking about, but specifically what we're going to be talking about is what kind of data sharing do you need to use AI in a good way? Oftentimes you hear, and not just in telecoms, and in fact, mostly not in telecoms, that the companies wherever they are, they're not actually getting as much as they feel they could or should from AI that is not really helping them as much as they could.

And there are many issues there.

It's a very complex thing, but one of them is that typically we operate in a multi-vendor environment where, especially with the production of AI, you need to share data and you need to share data in a way that it's not just nominal.

You can actually understand that data, use it.

And this is not a new problem.

This is something we had even before, but with AI, you need more data.

So the problem becomes more urgent and more fundamental.

Hello everybody and welcome to Sparing Partners.

Today we're again talking about, but specifically what we're going to be talking about is what kind of data sharing do you need to use AI in a good way? Oftentimes you hear, and not just in telecoms, and in fact, mostly not in telecoms, that the companies wherever they are, they're not actually getting as much as they feel they could or should from AI that is not really helping them as much as they could.

And there are many issues there.

It's a very complex thing, but one of them is that typically we operate in a multi-vendor environment where, especially with the production of AI, you need to share data and you need to share data in a way that it's not just nominal.

You can actually understand that data, use it.

And this is not a new problem.

This is something we had even before, but with AI, you need more data.

So the problem becomes more urgent and more fundamental.