A highlight from Some foresight about the future of foresight

Future Tense
|

Automatic TRANSCRIPT

This is an abc podcast. The future is timeless and time consuming. Human pursuit hello. Welcome to future tense. Getting a prediction wrong can have catastrophic consequences if you're talking about geopolitics military strategy and it can be economically costly as well take fashion for example. It's one of the world's largest industries and council round two percent global gdp which is not to be sneezed at. It has always been really important brands to don't miss out on trends. And there's a lot of work. Going into mapping what is coming and there's always been these experts working with fashion trend. Forecasting camilla green time larson a researcher and consultant based in norway. In the seventy s. the first agency agencies started to pop up in paris and they were predicting the next and fashion. Then all of the big brands a have been used to paying high prices to get these insights from you. It's not a coincidence. If during a fashion week a lot of brands showcase similar looks. It's because they have been given kind of the same predictions from the agencies. So if agencies are presenting focused saying that you know animal patterns will be a big thing then. You would see several browns doing that. And it's not a coincidence that they all do seabra pattern at the same season so it has always been a big part of it but the big change now. Is that before. There used to be fashioned. People fashion experts fashion girls people with a lot of trend knowledge historical knowledge about how trends working and meeting people travelling going to trend hunt on the ground visiting cities going to milan looking at what women are wearing in europe and then reporting back and always trying to be the first to know so the big difference now is that with big tech big data. I they have this advantage that they have a lot of data and then they can use that data to forecast trends presumably more accurate than those fashion experts which brings us to the subject of today's program the changing role humans in the business of forecasting and problem solving now. I said changing because this isn't a simple narrative about how a is taking over the world. It's much more complicated and nuanced. We are still in a really face of exploring the potential of a fashion. So we don't know yet what it could do for fashion. One thing is to use to analyze your customer data and predict what product you should make more of what you should stop making. Because it doesn't sell. That's one thing. But and i can also be used to sign garments and this is that facebook has been trying out. Today did experiment with a i to sign a thousand fashion garments and then they reviewed what these comments looked like and most of them were kind of okay but it was funny because the i designed pants with three legs so that was kind of. Yeah something that's the machines still. They don't have common sense and they they just take garments and put them together in new ways to be creative but they don't understand that humans don't have three legs so it's still a way to go before signing with a is where it should be are. Yeah what i wanted to be. So the first point to make is that while artificial intelligence is being used more and more in the practice of prediction human creativity still has its place and that's true for political science as it is fashion or any other endeavor. It could be that the growing use of ai. Forecasting tools is simply part of our modern obsession with darter and machine learning. But may also reflect. Lingering doubts about the overall effectiveness of human analysis experts. Get a bit of a bum rap from my earlier work on expert political judgment. It's not that all experts performed as poorly as dr doing chimps. There were conditions under which some experts could perform significantly better than chance. There were conditions under which experts could perform at chance levels and there are even a few conditions under which experts did worse than chance. So it's a little more nuanced but experts are not equipped themselves all that well in that earlier. Work university of pennsylvania professor philip deadlock one of the leaders in forecasting theory and practice. We all have rather than facial need. Don't wait to figure out what's around the corner what what lurks in the future.

Coming up next