AI created content is seeping into every part of life, from court reports to medical advice and election data, it is becoming harder to distinguish the real-life content from the AI generated.
A recent news report exposed how an ‘expert’ brought in to give evidence in a US trial used ChatGPT to create his responses and produce evidence reports.
Prompts showed the engineering specialist directly asking AI to form reports defending the company’s use of particular procedures and to show how they were not at fault. Supplying ChatGPT with hundreds of court documents the individual even asked AI to show why he should be seen as an expert in his field and trusted.
The case raises questions around the use of AI in delivering expertise not only for court documents but in media reports and company content.
Should ‘expert’ pieces be reserved for genuine knowledge and opinion or are chatbots the inevitable next step?
We look at the rise in AI to generate ‘expert’ commentary, the flaws and how businesses can make sure they maintain their reputation as leaders in their field.
PR has always relied on experts to provide the technical detail behind press releases and content. Teams in the field know their industry and projects far better than any PR professional can. We turn content in to the right format, but the information itself comes from the individuals that work and know the industry.
As AI chatbots and search platforms become more widely used, I have noticed a growing trend in generic responses when asking teams for information to pull together expert features.
AI platforms are filtering in and becoming the new go to for responses. Whether individuals are running out of time or confidence to generate their own ideas they are increasingly relying on the instant answers generated from ready-made chatbot opinions.
However, there are some real risks in relying solely on AI responses.
Ease of instant responses can lead to inaccuracies
Taking just minutes for opinions to form, responses can be drafted around any topic with the right prompt. However, that is one of the main flaws. Content relies on asking the right questions in the right way.
It doesn’t consider the accuracy of the data being presented and has even been proven to make up statistics to fit its argument. If businesses rely on the data as fact, they risk errors creeping into messaging.
Investigations recently found that AI services made huge errors before the Scottish Election making up fake scandals, inventing candidates or giving incorrect dates. Think tank Demo found that AI bots gave voters misinformation to 34% of questions posed. With a fifth of voters using AI chatbots or search tools to get information about the parliamentary elections it left many with incorrect information before the election.
A recent report in Yahoo also found that half of AI health answers are wrong even if they do sound convincing. The claims being given are often unfounded, the footnotes lead nowhere and the chatbot is unable to ask further questions to see whether the question itself might be the wrong one to ask in the situation.
The worrying thing is that re-using inaccurate data can leave companies liable and nothing fuels a PR disaster better than false information and misleading opinions.
Incorrect assumptions lead to bias
Not only can the content have factual errors, but they can also form inherit biases. Algorithms are based on specific data that they are trained on. This means opinions are formed from the data they receive which can result in discriminatory and one-sided opinions that wouldn’t form in human interactions.
In 2019, a widely used algorithm in US hospitals designed to predict which patients would need extra medical care favoured white patients over black patients. Although race was not explicitly included, the algorithm relied on healthcare cost history, which correlated with race. Black patients often incurred lower healthcare costs despite similar medical needs, leading the platform to underestimate their care requirements.
AI can only copy opinion it can’t create it
AI is only as accurate as the data it is fed. It uses complex algorithms and data sets to make assumptions and pull together responses.
This is why AI content follows the same style and generic responses. They miss the natural flair of human content because AI tools recombine patterns from vast data sets to provide predictable responses and outputs. Using repetitive algorithms, the content created follows the same templates. This is completely against the requirements of thought leadership pieces that should generate new thinking and in-depth analysis of current trends and knowledge.
So how can leaders maintain their expert status?
It is important that companies embrace their knowledge and aren’t afraid to share it.
Keep innovating
New ideas are the key to innovation and something AI cannot support. The likes of Chat GPT pull together opinions from existing content and are only as knowledgeable as the information feeding them. It means that any response will be generic and not specific to your business.
Confidence in opinions shouldn’t be eroded
Don’t be afraid to have an opinion. Generic AI responses rarely cover topics in-depth and know a company as well as the people working in the sector day after day. Have the confidence to form your own opinions and write from experience. How does that concept affect your business? What have you learnt? Only then can advice and knowledge be truly useful.
Knowledge share is key to growth
Keep learning. Don’t rely on AI for knowledge. Keep testing and gaining expertise from reputable sources. Network with like-minded individuals to share experiences and best practise, complete industry training and read from trusted sources.
AI is a useful tool in generating ideas, but it shouldn’t be used to replace human thought and knowledge.
The world doesn’t need more generic content it needs strong thought leaders to stand up and share their experiences.
If you are struggling to find your voice and share your experience then why not get in touch? Creative Word PR can help to generate non-AI created content which speaks directly to your audience with key messaging relevant to your market.
Find out more by contacting cheryl@creativewordpr.co.uk
Creative Word PR 