As part of a months-long project, AP worked with Al Jazeera and technology companies Vidrovr, Metaliquid and Limecraft to explore how artificial intelligence and machine learning can fully automate the process of producing raw and edited content shot lists. IBC’s Media Innovation Accelerator program facilitates collaborations to address complex media and entertainment business and technology challenges.
“This is a big, giant leap forward,” MacIntyre said Wednesday during an IBC webinar.
The objective, MacIntyre said, is for automation to transition a manual and laborious process to one where video journalists and producers can check an AI-created shot list and spend more time on the actual newsgathering process, allowing journalists to report out more stories and cover more news.
MacIntyre detailed the inception and promise of the project:
The biggest thing that we have experienced in recent years is just the fact that there is now more and more video in the world. We are dealing with, even inside our agency, probably 15 to 20,000 hours of live video that we are transmitting a year. That is hundreds of thousands of edited clips and we are not alone.We also know the news media industry is very challenged right now in terms of resources. The economics have changed massively, and what we need to do is emancipate people’s time to do creative things. So being able to automate and being able to use AI in this way, to do something that is probably the biggest time suck, the biggest manual process that we do in terms of manually shot listing and manually transcribing soundbites – you can imagine how many interviews the AP does every year – that really lies at the heart of this and I think we took a step forward here.
AP on Thursday also demonstrated findings from another project in collaboration with Al Jazeera, Reuters, BBC and RTE, exploring how AI can help news organizations moderate live content, allowing them to both vet video content and integrate automated content moderation into live production workflows.