Benjamin Beavan, Senior Global Development Manager at Getty Images

I'm Ben. I'm the Global Development Manager for Getty Images. And I work on our new AI products and how we take those to market, and also with our biggest corporate customers, and how we help them as a partner. Sure. So today, on stage, I talked about how AI can help us make creative decisions smarter. So a lot of decisions are made by marketers and creatives, mostly opinion-based. And now we have the ability to look at big data sets in terms of which visuals, which hashtags, which textual elements, et cetera, et cetera. And there's lots of variables performed. And then use that data to help inform the creative decision-making process to make it stronger. So for us ourselves, that's kind of different, but mostly I've been talking about how we help our clients. So we work with a partner called Cortex. What that allows us to do is to look at the data collected on social media, so engagement data, so likes, comments, shares, retweets, et cetera, and they're different across the different platforms, to take that data back and then show which elements had an effect on performance. And so that data really allows our clients to make smarter decisions when they're choosing pictures, when they're building their strategy. And that can either be delivered in a static way or in real time through a platform. At the moment, we mostly help our clients with stills through this platform, but that's evolving to video, as video is an important part of how marketers communicate. You know, engagement rates are really, really important, and actually video is one of the best ways to engage customers and to keep them interested in something. And so what you can do is you can break down videos into frame by frame and then start to analyze what's happening within those in the same way as pictures and then deliver those insights in a different way. One of the complications of that is that video may have various elements, so in this case it'd be fairly simple, you know, it's an interview, there's two people, it's not going to change, but where a video would contain different elements every few seconds, it becomes harder to track that. But what we're really trying to do is understand how a visual narrative can be created to help our customers communicate in a stronger way, to build awareness of their brand, to build engagement, and ultimately drive sales of what their product is. At the moment, as I said, you know, those decisions are mostly creative, so it could be the loudest voice in the room, it could be the most senior person in the room, people are going to have different opinions, so one of the questions I asked in there was, you know, for a holiday brand, are pictures of the city or pictures of the beach going to perform better? And it was pretty much a 50-50 split, and actually it was the pictures of the city that significantly performed better than those of the beach, which was quite surprising for some of those in the room. And so that's an example of where you can use the data to help make the creative decision-making process better. Yeah, I think AI now, with computers, can see pictures in a way that humans would take a long time to do, so they can just look at patterns, so you know, you can just say male, female, long hair, short, well, medium-length hair, beard, no beard, and so that's really helpful in spotting patterns. So if you look at really big data sets, you may have like 100 million images, and you might want to say, okay, let's just start to categorize those quickly, and the human would take a long time to do that, so you can add metadata using tags, and then from understanding what's in the picture, whether it be color, whether it be subject, whether it be male, female, all those kinds of points, start to understand the patterns. And so that's what I was really talking about today, is how can you take that capability, apply it to things that are already taking place, and then optimize those, so how can you drive engagement higher as a result of using that data? So the data itself is really the fuel, and then the optimization is really where they're driving to.