'I'm being paid to fix issues caused by AI'

Sarah Skidd is a product marketing manager based in Arizona who works with technology and start-up companies. Recently, she was contacted by a content agency to urgently fix website copy for a hospitality client. The original text had been created using generative AI, but instead of saving money, it caused many problems. Sarah described the AI-generated writing as very basic and uninteresting. It was supposed to attract customers, but it ended up being dull and generic. To fix the problem, Sarah spent about 20 hours rewriting the entire website, charging $100 per hour for her work. She did not just make small edits; she had to completely redo the content. Sarah is not worried about AI taking over her job. She believes that skilled writers will always be needed, even as more businesses use AI tools like ChatGPT. She has heard from other writers who say that most of their work now involves fixing AI-generated content. Sarah is not against AI. In fact, she sees its benefits. Her husband and son are dyslexic, and for them, AI tools can make writing much easier and even life-changing.

In recent years, generative AI has become more popular, and many businesses are using tools like ChatGPT and Google Gemini to save time and money. According to a survey, more than a third of small businesses plan to use more AI in the next two years, and this number is even higher for companies that want to grow quickly. However, using AI is not always as easy as it seems. Sophie Warner, who co-owns a digital marketing agency in the UK, has noticed a big increase in clients asking for help after trying to use AI on their own. In the past, clients would contact her if they had problems with their website or wanted to add new features. Now, many try to use ChatGPT first. Some clients have added code to their websites based on AI suggestions, which has led to websites crashing or becoming vulnerable to hackers. Sophie shared an example of a client who tried to update their event page using ChatGPT. What should have been a simple 15-minute task ended up costing the client about £360 and left their website down for three days. This problem is not limited to small businesses; larger companies are also making similar mistakes. Sophie and her team now spend more time educating clients about the risks of using AI. They often have to charge investigation fees to figure out what went wrong, and fixing these mistakes usually takes longer than if professionals had been involved from the start.

Professor Feng Li from Bayes Business School says that some businesses are too optimistic about what AI can do. AI is known to make mistakes, sometimes creating content that is irrelevant or even made up. He stresses the importance of human oversight. Poor use of AI can damage a company’s reputation and lead to unexpected costs, often requiring professionals to fix the problems. In India, copywriter Kashish Barot edits AI-generated content for US clients to make it sound more natural and less like it was written by a computer. She says that while AI is fast, good editing takes time because it requires careful thought and understanding of language. Many companies are experimenting with AI without clear goals or the right infrastructure. Professor Li warns that relying on AI tools without understanding their limitations can lead to poor results. OpenAI, the company behind ChatGPT, says that the quality of results depends on the model used, the user’s experience, and how the instructions are given. There are different versions of ChatGPT, each with its own strengths. Sophie Warner is both concerned and hopeful about the future of AI. She says that while AI can be a quick and cheap solution, it often fails to capture a brand’s unique identity and can make content look generic. In her view, AI is a useful tool, but it cannot replace the value of human expertise and understanding in marketing and communication.