Many of those using Google’s search engine recently have commented on how accurate the new artificial intelligence (AI) response can be, and how much quicker it is at getting answers to simple questions. However, a number of users have also commented that it is by no means perfect and still requires further drilling down through search results to verify its findings. In last month’s issue we spoke to Nigel Toon about his book, How AI Thinks, and he pointed to the enormous benefits we will receive from artificial intelligence. This month I was struck by an article on page 69 about how many UK doctors are using AI tools, such as ChatGPT for generating documentation after patient appointments, as well as to assist with differential diagnosis. The lead researcher, Dr Charlotte Blease, who has published more than 150 articles in peer-reviewed journals and authored The Nocebo Effect: When Words Make You Sick, highlighted the huge potential of AI but also warned that the potential for its tools to introduce errors, exacerbate biases, and compromise patient privacy is significant. She says that robust guidelines and training programmes to ensure their safe and effective use need to be implemented urgently. However, news of the benefits of AI seem endless. Avon and Somerset police were recently reported to be trialling an Australian AI tool that can analyse video footage, financial transactions, social media, emails and other documents simultaneously. In one task it took just 30 hours to review information on a crime that would previously have taken 81 years to sift through. This is an enormous leap forward. I don’t expect them to be able to find out who nicked my favouirite fountain pen in sixth form, but the possibilities for families who have suffered loss and had no justice could be much improved.
