I just finished reading an article on Forbes titled “The Jobs That Will Fall First As AI Takes Over the Workplace.” The article referenced reports from PWC, McKinsey, and the World Economic Forum that estimate that 60 per cent of current jobs will require significant adaptation due to AI.
The report further identified skills and professions like graphic design, copywriting, and basic journalism as vulnerable to disruption from tools like DALL-E and GPT-derived platforms, which produce content at scale. It further referenced a 2024 Pew Research Centre report that predicts 30 per cent of media jobs could be automated by 2035.
|
|
Similarly, in a number of panel discussions and conversations on AI, especially with journalists and writers, the dominant narrative was that artificial intelligence would replace writers, researchers, and media professionals. At least since the beginning of 2022 or thereabouts, this has been the case. Many of these have led to apprehensions and fears about embracing the technology, as it is a potential source of job loss in the immediate future.
In the race toward automation, it was easy to assume that those who trade in words, ideas, and storytelling would be left behind. But as the AI revolution unfolds in real time, something quite different is becoming clear: writing, media, and research are not only surviving the AI boom, they are powering it.
|
|
|
|---|
In the United States, the Writers Association once protested against the adoption of Openai’s Chatgpt, arguing that it could take their jobs.
Writing is the new coding
Every time you prompt a chatbot through Gemini, ChatGPT, Perplexity, or any of these generative AI models, whether you are generating a script or asking an AI model to draft an email for you, what you are engaging in is a form of writing. Prompt engineering, once a fringe curiosity, is now a full-time profession. And what is it? The art of clear, precise communication with machines – the daily art of giving instructions/prompts for generative AI models to get desirable results. It is a general belief that your results from AI are as good as the instructions, examples, and context you feed it. All of these activities are powered by writing. This means you cannot get the best out of generative AI if your prompt engineering game is not at an elite level.
The reality of engaging with generative AI in today’s world is that AIs live and breathe through text. The better your language, the better your outcome. In this new era, writers don’t become obsolete; they become infrastructure.
AI doesn’t just generate content; it shapes what we see, share, and believe. Newsrooms are adopting AI to assist with reporting, but they are also on the frontlines of a new wave of challenges: deepfakes, disinformation, algorithmic bias, and AI-generated propaganda.
Media isn’t just covering the AI story; media is the story. It’s where AI’s societal, ethical, and political implications are tested in real time.
This makes journalists, fact-checkers, content strategists, and editors more essential than ever. They bring context, nuance, and credibility to the equation, things no model can mimic without human oversight.
Today, when almost all knowledge producers are wary of AI-generated texts, human editors are needed more than ever to gatekeep the process. In an era where generating deepfakes and potential disinformation can be done with a single prompt, fact-checkers are now very important experts that we need in the efforts to sanitise the information ecosystem.
Research as the compass in the chaos
Behind every major AI model is a trove of research. From machine learning breakthroughs to governance frameworks and impact assessments, the research ecosystem is the foundation of AI’s progress and accountability. But beyond technical research, we need interdisciplinary lenses: social scientists, policy thinkers, ethnographers, and development experts who ask the hard questions about power, equity, access, and impact.
As someone working at the intersection of media, governance and digital technologies, I have seen how essential this research is, not just in boardrooms and labs, but in local newsrooms, policy roundtables, and grassroots communities across Africa.
Evidence-based journalism as the future
If you write, tell stories, shape narratives, or explore complex questions, you are not on the sidelines of the AI boom; you’re holding the map, the pen, and the mirror.
This is the time for investigative journalists, storytellers, media innovators, fact-checkers and curious researchers to shape how AI is understood, governed, and deployed.
Technology only becomes meaningful when interpreted, contextualised, and made accessible by humans who care about truth, equity, and impact.
READ ALSO: Degrees, AI anchors, and the future of journalism: Why skills and ethics must lead the way, By Akintunde Babatunde
If you cast your mind back to the early days of chatgpt by openai and the foundation of its model and how its data were trained, even though OpenAI has not publicly shared the precise percentage of journalism or writing content in ChatGPT’s training data, estimations put the training data composition for GPT-3 and GPT-4) to be 60% Common Crawl, which includes a wide range of internet content like news articles, blogs, and online forums; 22% WebText2 curated quality web pages, which likely include journalism and essays; about 6% literary and non-fiction books and some journalistic nonfiction; and about 3% encyclopedia-style content. Little wonder at the improvement we have seen on these models where links to news articles are now embedded in search results as a result of content licensing partnerships that these AI companies are now signing with media organisations, as in the case of the Financial Times, which allows OpenAI to use its content for training purposes; Associated Press and Axel Springer which provides access to their archives.
All of these are examples that point to the centrality of writing and journalism in the AI era and how smart, original writers will not be left behind. Still, they will also shape the future of generative AI because in an age where AI models are at risk of hallucinations and bias, we need training data that is true, evidence-based, and accurate. Unlike journalism and academic research, no other writing process comes close to that. So, without journalistic-quality inputs, trustworthy, fact-checked writings done by humans, AI risks becoming an amplifier of misinformation, which will put its credibility at risk.
As we continue to embrace AI and media organisations license their archives to AI companies (like Financial Times and OpenAI and Axel Springer and OpenAI), journalism becomes a premium data asset. This is especially important in regions like Africa, where journalism, research, and civic technology will shape how AI serves development, accountability, and public good.
In the age of generative AI, journalism and research will not just survive, they will shape the very credibility and future of intelligent systems
![Ekiti state on map. [Photo credit: Wikipedia]](https://i0.wp.com/media.premiumtimesng.com/wp-content/files/2017/12/Nigeria_-_Ekiti.svg_-e1514118425834.png?fit=1000%2C769&ssl=1)
