ChatGPT is a new artificial intelligence chatbot that is becoming increasingly popular amongst all sectors, sparking a lot of interest in the power of AI and its implications. AI technology is being used in manufacturing processes, in the US court system, facial recognition systems and even for sifting through CVs. But one of the latest trends is for AI to manage interactions with humans, resulting in a huge need for nuance and context, to reach the ideal response.
Seán Moncrieff, a well-known Irish broadcaster, writer and media personality, recently set ChatGPT the challenge of writing his Irish Times column in the same style as his own. While it was able to drop a few interesting comments about the future of AI, it is clear that the drivel it produced was banal and cliche-ridden and lacked the human awareness of Seán Moncrieff’s real columns. It especially highlighted the issue of bias in the computer age, as it absorbed the biases of its programmers and missed out crucial information.
This can be seen in many forms, from the customer service chatbots that pepper websites and prompt us without actually doing anything to help, to the facial recognition software that is flawed with racial and gender biases. There is also the fear of AI taking control in a ‘Skynet’ situation, running out of control from the programming that it has been given. This ensures that there is an inherent danger for AI, as it takes small decisions that are easy to brush off but that can have terrible long-term repercussions e.g Napalm.
As AI continues to be unleashed in to the world, it is essential to maintain a certain level of human oversight and to keep this technology held in check. AI poses an existential threat, and the window for gaining control over it is small. As the debate on how to manage AI technology continues, Seán Moncrieff’s experiment showed us that an AI chatbot isn’t the great all-powerful supercomputer people often think it is. For now, we must stay open and curious as we navigate this brave new world of AI.
