Artificial intelligence (AI) is set to become part of the boardroom agenda as organisations look at how they can shave off more from operational costs and work out how they can grow the business during these uncertain times.
Among the areas of AI development that has caught the public’s imagination is ChatGPT, a new generation of conversational AI model that promises human-like conversations. The Turing Test is among the early tests of true AI and it seems not only is ChatGPT able to fool humans into thinking they are having a conversation with a real person, it can also write convincing reports, articles, essays and even novels.
The introductory paragraph from a recent McKinsey article, for instance, was convincingly generated by ChatGPT. While the authors, unsurprisingly, say, it’s not as good as a real writer, they had to concede how effective the AI’s effort was. In fact ChatGPT is so convincing that if someone who does not work full time at one of the big consulting firms was presented with such an AI-generated paper, they could easily think it was written by a human.
Should we be concerned?
By trawling all the publicly available information available on the internet and building an understanding of meaningful human conversations, there is no doubt generative AI systems like ChatGPT, will soon learn to mimic human writers extremely well. Use cases include producing highly readable operations manuals, how-to guides, safety documents and summarising legal reports.
Research from Finnish security firm, Withsense, supported by the CC-Driver project within the European Union’s Horizon 2020 programme, warns that this technology will almost certainly be exploited by criminals. “The generation of versatile natural language text from a small amount of input will inevitably interest criminals, especially cyber criminals – if it hasn’t already. Likewise, anyone who uses the web to spread scams, fake news or misinformation in general may have an interest in a tool that creates credible, possibly even compelling, text at superhuman speeds,” the researchers warned.
Going forward, AI is on a trajectory to revolutionise many of the “manual” tasks people do during work. And yes this includes those jobs involving writing, summarising reports and tasks like coding, which, at its most primitive level, involves joining together existing algorithms in order to achieve something meaningful. Many will argue that as humans, we are capable of having original ideas. But like algorithms, great ideas are built on the shoulders of giants. It’s not beyond comprehension that an AI could mimic this by drawing on existing published material.
Undoubtedly, with the wealth information that can easily be gleaned from the internet, an AI has the potential to run organisations better than their existing CEOs. Now that is something that’s unlikely to be considered in the boardroom.