Is AI a good thing or a bad thing? There seems to be a lot of debate at the moment. But perhaps that is a stupid question, as no ethical issues are quite so simple and binary. If there was a right or wrong answer, we wouldn’t need to discuss it. So, I decided to ask ChatGPT the question, “Is AI a good thing or a bad thing?” This is what it replied:

“As an AI language model, I don’t have personal opinions or beliefs. However, AI technology has the potential to do a lot of good, such as improving medical care, increasing efficiency and accuracy in industries, and facilitating scientific research. At the same time, like all technologies, AI can be misused, and there are concerns about things like bias in algorithms, job displacement, and privacy and security issues. Ultimately, it is up to human beings to use AI responsibly and to ensure that its benefits are shared fairly.”

It is a splendidly balanced, but one might say evasive, reply. It begins by saying that “I don’t have personal opinions or beliefs,” then goes on to give its opinion that “AI technology has a lot of potential to do good” while at the same time “it can be misused”. (We’ll come back to the use of ‘I’ and ‘personal’ later.)

Clearly it does have an opinion; its opinion is that AI has the potential to do a lot of good and that it may be misused. That is so banal as to be meaningless. It could have been said about every new technology from the invention of the printing press in the 15th century, to the Jacquard Loom in 1801, to the television or mobile phone. So, I asked ChatGPT another question: “Why is AI a good thing?” Its answer was much as before: “AI has the potential to greatly enhance our productivity, efficiency and overall quality of life.”

When I told it that I disagreed its rejoinder was that “Overall my purpose as an AI assistant is to be helpful, reliable and ethical and my developers take adequate measures to ensure that I operate in a fair and responsible manner.”

I must admit I am impressed by its ability to engage with sequential questions as though it was having “a constructive discussion” (its phrase), but it is not really a conversation because each answer is independent of the previous one. Currently, it doesn’t engage with the questioner’s thought process, only with its own database – one question or comment at a time.

Of course, chatbots are only a subset of AI and ChatGPT only one version of a chatbot. It is certainly a great improvement on the bot which fails to answer my questions about my bank account and is fronted by a cartoon face called Ruby the Robot. ChatGPT will be right in its prediction, I expect, that AI will support improvements to health care and research and will increase efficiency in some transactions while replacing some people’s jobs. What It didn’t tell me is that it will also provide information which is incorrect, incomplete, or misleading and that human brains will have to continue to make judgments as to its value.

It is the use of ‘I’ in ChatGPT’s answers that bothers me. It is not a person. It tells me it has “no personal opinions or beliefs.” Correct. But it claims a personal identity in the use of ‘I’ and its designers have deliberately chosen to imply by the use of ‘I’ that it is an intentional, unique entity rather than a massive interactive database. That illusion of individuality rather contradicts, to my mind, ChatGPT’s statement that its “developers take adequate measures to make sure I (sic) operate in a fair and responsible manner.” It is not fair and responsible to mislead the user into thinking it is engaging with a thinking individual rather than a lifeless machine.

In the end, we must agree with ChatGPT (which claims to have no opinions for us to agree with) that “It is up to human beings to use AI responsibly and ensure the benefits are shared fairly.” This sounds like the amorality of the technician; I just make this stuff. It’s up to you how you use it.

That dilemma was why I asked the question. It requires a very thoughtful answer – which I didn’t get. I didn’t get it because humans don’t know the answer and the chatbot only regurgitates what human beings have told it. Is AI a good thing or a bad thing? I still don’t know. Do you?

Harry Cayton is a sought-after global authority on regulatory practices who created the Professional Standards Authority (PSA) and pioneered right-touch regulation. He is a regular Ascend Voices contributor.