The past year has been full of excitement about the opportunities presented by AI, but also trepidation that AI will cost media industry jobs.

The burning question remains: What is the future of humans in media amidst the AI revolution?

We recently spoke with four media industry leaders about AI and other hot topics going into 2024. One of the key themes that emerged was the importance of incorporating human expertise into AI initiatives to maximise success.

Why the human touch is important for AI

“There are huge risks in getting a bot to write content for you that you just automatically publish to your site,” said Daisy Donald, principal consultant and head of Americas at FT Strategies. “I just don’t think it’s wise to ever fully replace a journalist with a bot because there’s just too much risk to your business unless your business model is a volume game and you can take the hit on quality sometimes. It just really depends on your brand.”

The brand and ethics risks posed by AI were top of mind among the leaders we spoke with. They emphasised the importance of high-quality, original content for engaging audiences and maintaining their trust. But they also acknowledged the potential for AI to make content producers’ jobs easier in the long run, and shared two examples of how publishers are already using AI tools combined with human expertise to create more and better content.

Example #1: content translation

Adley Bowden, head of individual investor and editor-in-chief at Morningstar Wealth, described how the company is using AI for content translation, which leans into the strengths of large language models.

“We’ve got a dozen languages, 20 different sites, and, for us, the trust and the authority of our content is really important. We can’t just do Google Translate and off we go,” Bowden said.

Morningstar is testing AI translation services, including options integrated within its CMS, which have cut hours out of the process of translating a story and made its newsroom more efficient. But the company also inserts humans into the translation process to check accuracy and supplement AI with some manual translation work.

“AI will allow news organisations with a global footprint to provide higher quality content across geographies. That opens up a whole new challenge of localising versus globalising. But it’s an area where we’ve had success with AI, and I encourage anyone who hasn’t tried it to look into it,” Bowden said.

Example #2: personalisation

Many publishers have started using AI-personalised home pages based on what visitors have read previously, said Madeleine White, head of international at Poool and editor-in-chief at the Audiencers. This AI-facilitated personalisation removes articles that a visitor has read or clicked on previously and presents new articles that match their interests.

“Every single publisher I’ve spoken to who is doing this has a maximum of maybe 70% of their home page that is personalised through AI,” White said. “The rest is still in the control of their editorial team to really push the stories that matter because a machine can’t know what’s going on in the local community right now and what everyone is talking about in offline conversations. It’s really important we still have that control as humans and you have a human nature to your publication.”

AI also can’t replace the humans who build and engage the real-world communities that become loyal audiences for publishers. White cited the example of Blick, a Swiss publisher that launched an innovative campaign to reach skiing, snowboarding, and hiking enthusiasts by sponsoring a competition for the best ski resort and providing interactive opportunities for visitors to register their vote.

“Looking back at any big technological change that has happened, like the Industrial Revolution, it didn’t put everyone out of a job. It actually helped create new jobs and opportunities,” said Birger Soiland, vice president of sales at Norkon. “And I think we’re at the cusp of something similar now where we’re going to see AI help us and generative AI being helpful to the newsroom.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .