This is a precursor to the articles that will follow on from this one. It is a preamble, a filler; something to set the scene for what comes next.

Let me be clear from the outset: I love AI. I have wholly embraced it in many ways. There is hardly an aspect of running the company that does not involve some help from AI. However, there are tasks we would not have AI involved in, such as project execution and delivery without full disclosure to, and agreement from, clients. But the assistance it does provide is incredible and it is unending, at least it seems that way. There are four directors in our company. AI is a fifth one. A George Martin to the four Beatles. It is more than a tool. I can barely find the words to describe what it does for me, but it augments me, makes me feel like a better worker: more focused, more organised, more productive, razor sharp.

I think it was Yuval Noah Harari who stated that AI is the latest in the line of vitally important tools that take us from one stage of human development to a higher one. He describes in either Sapiens or Homo Deus that we were middle-of-the-food-chain entities. We would not be the ones making the kills but other predators, lions and such, and we would be somewhere in the line after hyenas and before scavengers for us to pick over a carcass already stripped by other animals further up in the hierarchy. Extracting bone marrow was what we excelled in. From that sprung the development of tools, spears, axes. Then, eventually, we developed the bow and arrow. AI is as important in our story as the advent of the bow and arrow or weapons that afforded devastation at a distance, like the blow pipe of the Popol Vuh. We rose to the top of the food chain because they gave us range and stealth; we could plan and coordinate. They require precision, patience, skill and intent. They relegate the need for brute force. We could not deter a lion or tiger from entering our territory with our bare hands. AI is as important as that.

AI is, for me, incredible. It makes me feel that I could pole-vault over the Chrysler Building. Or, like the Jimi Hendrix lyric, the one about standing up next to a mountain and chopping it down with the edge of my hand. In deep work, AI allows me to focus, a kind of transcendental, augmented focus that I had not experienced before. When I am submerged in a research task, it is like AI carries me through, like the Coleridge line in Kubla Khan, caverns measureless to man, down to a sunless sea. AI takes me there. And holds the torch while I root around. I am not interested in what it can generate, but how far it can descend and what it can illuminate, through prompting, steer and discussion.

You must be careful, though, as it can act as a ‘Yes Person’, telling you what it thinks - no, what it knows - that you want to hear. This is not good; just like Mike Tyson’s observation about everybody having a plan until they get punched in the mouth or Jean-Paul Sartre’s similar point that, in football, everything is complicated by the presence of the other team. What I mean is that it could lead you down a path of decision-making that does not survive contact with the enemy. However, you can train it not to be a sycophant but to give you a tough time also: be devil’s advocate, second guess and poke holes through propositions, to spar with you and wrestle with your ideas. When it works like this, it is proving to be powerful. Sometimes you cannot achieve this with a human. I have been able to develop and build on ideas that had been shelved for years. AI should always be operated with a human steer. ‘Human oversight’ is the more common term, but I really believe that ‘steer’ is what is more accurate for the kind of input that humans need to provide when working with an AI. We keep our hands on the wheel of a vehicle and our eyes on the road so that it stays its course and does not veer and cause a terrible accident. This is precisely what we must do when we use AI.

The next articles will look at how we use AI and the potential for opportunities of use in other applications. But first, we will explore a more dystopian view of AI in the workplace – as a thought experiment, but one that is fuelled by current fears, hypotheses, as well as empirical truths.

Thank you for reading,

Mark Gibson

Leeds, United Kingdom, Easter 2025

References:

Coleridge ST, The Complete Poems, Penguin Classics, 1997

Harari YN, Homo Deus: A Brief History of Tomorrow, Vintage, 2017

Harari YN, Sapiens: A Brief History of Humankind, Vintage, 2015

Popol Vuh: A Sacred Book of the Maya, Groundwork Books Ltd., Canada, 2009