How we use AI in our reporting work
Content Director
Harnessing AI's potential in our publications – with an eye on its pitfalls
Life has felt a little bit different ever since the demo of what was described as “a new artificial intelligence chatbot” was launched in late 2022. A couple of years on, that unfamiliar phrase now sounds completely normal, while AI and the groundbreaking ChatGPT tool have become part of our working world and – without wishing to sound too robotic – the way we operate. Barely a day seems to pass without murmurings of a new game-changing platform or sections of the media moving the finger of blame on to the bots for “stealing our jobs”.
Yet while many sectors have seen significant and rapid revolution, the publishing world has often had to search a bit harder to find its seat on the AI rollercoaster. Yes, there have been plenty of technological advances that have quickly become common practice in the workplace, particularly among our digital and social teams, but – when it comes to producing reports – at times it feels like we are still waiting for the smoking gun that has transformed other industries.
The top of the hype curve
At September’s Future of Media Technology Conference in London, it was clear the mood was changing, some of the initial hype replaced with a more conservative outlook. “We heard suggestions that as AI is so far mostly limited to efficiency gains, it hasn’t lived up to its hype and that we’ve reached the top of the hype curve and we’re entering the trough of disillusionment,” said Dickon Ross, former Chair of the British Society of Magazine Editors.
Whether we’re witnessing overhype or a complete overhaul, at Touchline we will continue to approach AI in the same way. We will never cut out the human touch in our work or allow AI to replace the superb skillset of our team, but at the same time we recognise its enormous potential. When used strategically, with caution, and with sensitivity about data and confidentiality, there are clear benefits for our work and our clients.
Strategic, thoughtful use
Our content team has always been quick to embrace the tools that help improve the efficiency of everyday language-based tasks, such as transcribing, translating, proofreading or writing headlines, and we are constantly on the lookout for new opportunities in the ever-evolving AI landscape.
Only the other week, a colleague recommended a platform that can compare documents side by side. A project that begins with a Word file and ends with a beautifully designed pdf has many twists and turns along the way that will never be seen by the target audience. These can include a whole raft of text changes from several stakeholders throughout the production process – some shared by email, some scribbled then scanned on a pdf, others hurriedly mentioned among the polite goodbyes on a Teams call.
So, to have a tool that is able to flag any differences between file versions and help you check everything has been implemented correctly, as well as highlight any discrepancies or accidentally repeated sentences, can be invaluable.
Has this new tool revolutionised the way we work? That would be a stretch. Does it save us time and allow us to deliver a better service for our clients? Absolutely. And more time saved on day-to-day tasks translates into more time for our team to get creative.
Authentic storytelling > generated copy
One area, however, where we will not use AI is writing copy. Remember all that talk about robots replacing the writers? Well, we’re not there yet, certainly not when it comes to creative, authentic writing (the allegations about one well-known sports media outlet publishing articles by non-existent writers with AI-generated headshots caused quite the furore).
Of course, we have done some extensive research into what’s available and will always keep an eye out for any new generative copywriting tools, but we strongly believe that stories should be crafted by humans. Telling authentic, original, trustworthy stories is at the heart of what we do as an agency and we feel that our editorial team delivers content that – for now, at any rate – is far superior to anything produced by a computer.
Hand-in-hand with a human
One thing we have found to be truly transformative in our reporting work has been the impact of AI on imagery, both in the moving and static sense. The capacity of tools such as Midjourney to create photos based purely on text prompts was an instant headline-grabber, but lately we have been exploring the more intricate details on offer around conceptual imagery. For instance, you can take this one step further by uploading an image you personally have sourced or provided by a client as a starting point, then use the platform to develop the shot as per your instructions. All of a sudden, your new image matches the style of the other photos used in a report or is placed in a setting that aligns with the relevant chapter theme.
It’s a similar story with illustrations and other graphics. We were recently tasked with creating a new set of illustrations for a publication. Using the original sketches, we explored how various AI tools could develop these into something more in line with the client’s wishes. Before the day was out, on the back of some carefully worded instructions, our creative team had an impressive shortlist of new drawings, each unique in its own way and all evolving from – but still remaining faithful to – the originals. These illustrations and images can now be easily animated too, something which allowed us to produce a video storyboard for a client in record time.
Transformative technology, transformative timelines, but in our experience the overall outcome remains a reflection of the designer’s talent – just as it always has been with traditional tools like Adobe Photoshop. AI platforms are only as good as the designer who is guiding them – a crucial factor in us continuing to deliver a premium product.
Looking ahead
As our design team embraces the fast-changing tech, so too do we see a growing interest from our clients, both in how we collaborate and how AI can help them. This year, we have designed an AI strategy for one of the world’s biggest sports organisations and have also created the Trading with Intelligence report for the World Trade Organization – in both instances using AI-generated imagery.
And how we use AI is no less important to us and our clients than what we’re using it for. Touchline’s Digital Art Director Edson Rosas believes the long-awaited launch of the Adobe Firefly generative video technology will be a game-changer in 2025.
“Many AI platforms use footage from across the internet to train their machine learning models, often pulling from various sources. This approach can lead to the use of others’ work without permission, potentially infringing on copyright or IP rights,” he explains.
“Adobe’s new Firefly technology, however, ensures that users – who are required to purchase a subscription for Adobe’s Creative Cloud – work exclusively with a system trained on resources and references sourced from Adobe-approved stock footage and imagery.”
This provides clients with confidence that we are employing AI in a responsible and ethical manner.
Whatever comes next – and one thing we know for sure is there will be plenty more developments – one key element will remain at the core of our work, as it was in all of the above examples. A human hand will play a pivotal part in creating the content and producing our reports, which is exactly how we want it to be at Touchline.
