Why I Am Not Afraid of Artificial Intelligence

Artificial intelligence seems to divide opinion. Depending on whom you speak to, it is either about to transform the world for the better or destroy it completely. Utopia or Armageddon? I have never found either stance particularly helpful, or particularly honest for that matter.

I have been using ChatGPT since about March 2023 and still find it a remarkable piece of technology.

The novelty has not worn off and if anything, the opposite has happened. The more I use it, the more ideas it generates and the more projects I discover that I simply would not have attempted before.

I have been fortunate enough to live through several technological revolutions. Personal computers arrived in our offices and homes. The internet changed how we found information. Mobile phones meant that we were always connected. Each changed the way we lived and worked, and each was greeted with equal measures of excitement and concern.

Every major technological breakthrough, from the printing press to the internet, has followed a similar path.

The printing press fuelled extraordinary advances in literacy and scientific progress, but it also enabled the rapid spread of propaganda, religious conflict and manuals on witch hunting. Radio and television informed the public and helped create a shared culture, but they also became a vehicle for state propaganda and mass manipulation. The internet gave us access to much of the world’s knowledge, yet it also gave rise to cyberbullying, grooming and radicalisation.

Artificial intelligence amplifies human nature on a scale we have never seen before and, regrettably, human nature is not always charitable or benevolent. I must therefore conclude that AI will be no different.

Viewed through this historical lens, the central danger is not that AI will suddenly acquire a mind of its own and decide to harm us. It is that self-interested people will use it to exploit and manipulate others far more effectively than they could before.

Technology is rarely inherently good or bad. What matters is what people choose to do with it.

When I first started using ChatGPT, I assume that, like most people, I asked it to write a few messages. The results, I am afraid, are now lost to history.

If that is all you use AI for, and who am I to judge? It does seem a little like using a bulldozer to tow a caravan. Yes, it can do it, but it can do so much more.

How many of us have a textbook sitting on a shelf, purchased with good intentions but never read? It might be about marketing, staff management or site safety. There is a very good chance that AI has already absorbed much of the knowledge contained within it and can help us begin to explore the subject.

There is a large group of people who believe that writing the perfect prompt is the key to receiving the perfect response. I have come to believe, perhaps a little cynically, that quite a few of them are also trying to sell us their book on the subject.

Carefully constructed prompts certainly have their place. If you are asking AI to write computer code or perform a precise technical task, I heartily recommend getting your ducks in a row before typing anything into that rectangular box.

For most of us running small businesses, there is no need to become a prompt engineer. AI has been presented as something we can address in plain English, and it will usually reply in plain English. We should take it at its word.

I now rarely give ChatGPT a single instruction and expect a finished answer. I use it as a very powerful thinking partner. The best ideas do not seem to arrive by asking better questions. I have usually found them by asking the next question. That is why, instead of calling it prompting, I call it a conversation.

A conversation accepts that the first question is merely the starting point, a stepping stone towards greater understanding. It also accepts that the destination may be discovered along the way.

I would never cede my judgement to AI. Accepting everything it says without question would be foolish.

I sometimes think of it as going out on patrol with someone riding shotgun. I am still responsible for where we go and the decisions we make, but it helps to have another pair of eyes scanning the landscape, questioning assumptions and occasionally spotting something I have missed.

Put another way, I am driving the conversation and remain responsible for where we end up. ChatGPT simply helps me arrive at the destination more safely, even when I did not know exactly where that destination was when we set off.

Over the last few years, I have used AI while developing health and safety procedures, standard operating procedures, staff guidance, website content, operational reports and business processes. None of these projects had a single correct answer. Most contained dozens of moving parts, each affecting the others.

What interested me was not whether AI could write these things. It was whether AI could help me think about them. That turned out to be a much more interesting question.

Nowadays, I rarely begin by asking for a finished document. More often, I simply describe the problem. We explore different approaches, question assumptions and consider what might happen if a process fails. Only then do we begin discussing the wording. We simplify. We start again. Eventually, almost without noticing, the document appears.

Looking back, I have learned something unexpected. The finished document is rarely the most valuable thing produced during those conversations. The thinking is.

I have probably learned as much from challenging AI as I have from accepting its suggestions. I treat it rather as I would an experienced colleague whose opinion I respect. I ask questions, challenge the answers and sometimes discover that I was wrong. Sometimes it is wrong. Either way, I usually come away understanding the problem better than when I started.

That, for me, has become the real benefit. Artificial intelligence has not replaced my experience. It has made better use of it. It has not removed the need for me to think; it has encouraged me to think more deeply.

Perhaps that is why I still find it remarkable. Not because it has all the answers, but because it keeps helping me ask the next question.

And what fascinates me most is what sensible, ordinary people choose to do with it.