Levelling the playing field: how JournalismAI preps smaller newsrooms for AI adoption

'The biggest challenge with AI is that it's unequal. The adoption of it is unequal, and that's where we come in. We're trying to level the playing field, because the bigger newsrooms are able to do it; they have the resources, the technical skills. The smaller newsrooms, not so much.'

Tshepo Tshabalalais Project Manager & Team Leader of JournalismAI, a research and training project run by Polis - the international journalism think-tank of the London School of Economics and Political Science.

A South African journalist whose work has appeared on radio, print and online, Tshabalala was also the web editor forJournalism.co.zaand founding digital editor ofJamlab Africa,both based at the University of the Witwatersrand. He also served as Jamlab's interim project director, then Director.

He shares learnings from Journalism AI's six-year exploration into AI, and a programme for small newsrooms considering AI adoption.

As a former journalist with a diverse academic background now at the forefront of AI in media, how do you describe yourself, and how do all of these disciplines converge for you?

I'm a stickler for learning. I just want to learn and know about everything and anything. And I think that's how my journey into AI came about. I'd describe myself as somebody that loves to travel and learn.

Having moved from a newsroom into media development prepared me for this space - and, like a lot of people, I came into the AI space not knowing much about it. I joined at the height - or the hype cycle - of AI. So I was learning as much as teaching, like everybody else. But I was really lucky to have a team that had the knowledge and experience needed prior to the hype of generative AI.

That's how sort of my background in journalism, my studies, financial journalism sort of converged to AI: I'm able to sort of pull in all those strengths, into the work that I do today.

The Journalism AI Academy has been offering its free online training course for small newsrooms for the past three and a half years; can you do a comparative of cohorts, in terms of reception and adoption, and their learnings?

Essentially, the comparison that I can make is the difference in participants from when they join the program to when they leave.

In our application form, we always ask about their AI literacy, and their hopes and fears.

At the end of the program, when we wrap up, we come back to those questions - and you can see the difference because with a better understanding of the technology, people are more open to try and get it into their news organisations, and perhaps talking to the wider organisation to say, "This is what I've learned; let's see what we can do."

And the beauty of that is the successes of the organisations that have been part of the Academy have implemented some form of AI strategy or tools, in their news organisations, which is always amazing to see.

Talk more about those implementations. What are you seeing in terms of adoption and process - where is the biggest uptake?

Our research from 2023 showed that AI adoption was mainly in news production, with some in news gathering and news distribution. The trajectory follows our first report in 2019 when the project started, which had higher figures for news production, then distribution and news gathering.

It really varies.Some people start slow, with an AI strategy first; some start with putting together guidelines to how they want to use AI in the newsrooms - and some people just allow the journalists to play around on the platforms and see what they can do.

There isn't a uniform AI adoption strategy or step for any newsroom - it all depends on resources and the technical expertise you have in the newsroom.

So, for example, a lot of the small newsrooms we work with either have the resources and don't have the technical expertise or the other way around or both - so some people come onto the program and never ever adopt AI in the newsroom because they just don't have the resources.

Is this common? If so, what are the major challenges for them?

The biggest challenge with AI is that it's unequal. The adoption of it is unequal, and that's where we come in. We're trying to level the playing field, because the bigger newsrooms are able to do it: they have the resources and the technical skills. The smaller newsrooms, not so much.

Either they are dealing with the technical and financial constraints of working in the newsroom, or they have other challenges - like dealing with the cultural change of new technology within the newsroom, which brings with it the fear of job losses.

We're seeing a repeat of history, with this introduction of new technology. Back when I was still in the newsroom and social media was coming to the fore, the older journalists didn't want to go into the field and record with the phone or take photos with the phone.

And we're seeing this with AI as well: in some of the newsrooms, there is that cultural shift challenge - abattle between younger people trying out the technology and managers or rest of the newsroom not wanting to implement or think about even using AI.

So the challenges vary depending on each organisation. They're not uniform, and we never try to make them uniform.

What advice would you give to the naysayers?

'The reality is that this technology has come out of the box, and there's no way of putting it back in. So, you need at least an awareness and an understanding of it, to know what it can do or what it cannot do. Because like any other technology, a lot of the people who get replaced in newsrooms are people who don't use or understand the technology by people who do and understand technology.'

We're not saying everybody must be a technical, AI-proficient person; we're saying have an understanding of it, know how it works, know what it can do, what it cannot do. In that way, you're able to see what or how you can work around it.

It's not going away. It will only get scarier and bigger if you avoid it. So, the advice we give to newsrooms is to play on these platforms. A lot of them have free versions. Play on them, understand what they do, and in that way, you would be able to embrace it a bit more.

I'm an optimist in the adoption of AI because we already use AI in a lot of things - from Google Maps to streaming and social media apps. I think what generative AI did was make AI user friendly - so over time, we could overcome this fear the same way back in the day, people overcame their fear of telephones and cell phones. Let's find what we did then; what did we learn then that we can apply now to AI?

What tools would you recommend for independent journalists or smaller newsrooms?

I wouldn't recommend any one because there are so many platforms out there, and AI is bigger than generative AI. So yes, it's good to play on generative AI platforms because those are the user-friendly models, but also be aware of your large language models, deep learning etc, and find courses that teach you the basics.

I find it difficult to recommend any platforms because you need to figure out what's useful for yourself - as do newsrooms, who have to figure out what's important for them; what problem they're trying to solve, and what tools and platforms to use.

What do you personally find most exciting, and or most daunting about this entire AI journey we are all now on?

The unknown. That is exciting. The unknown that at every moment you think you know something, tomorrow you don't know. And that's exciting, that I'm continuously learning, and that's what I enjoy, continuously learning.

Lucinda Jordaan

[email protected]