Musing upon Artificial Intelligence

Before AI became the mega trend of the 2020’s, I was already deep into the mathematics and code that makes it tick. It got me back into programming as a hobby and being a closet geek, I even enjoy the linear algebra!

Despite having presented about it at public and businesses events, I’ve never posted on the subject before, maybe refraining because there is just so much information out there; what could I possibly add?

This week, I have been catalysed to do so by our expert team at Ebbon-intelligence who have just published a very interesting article about AI and its uptake in the arena of vehicle sales in both the U.S. and U.K. markets: https://lnkd.in/e3swNWVi

A clear take away from their report is that it’s all about how you use it; if committed people use AI technologies in a smart way, then it can genuinely enhance your business. However, if your business expects AI to do it all, then it’s not going to be the Holy Grail.

I use the religious metaphor deliberately. Having spent thousands of hours listening to tech leaders - from those I deeply admire, like Demis Hassabis, to a recently growing chorus of AI evangelists, I've started to notice legitimate enthusiasm morph into something more concerning. In this era where share prices swing on tweets, AI hype is starting to birth its own sect of true believers.

My grandfather, who had a deep interest in Bible studies, used to tell me Old Testament stories, such as the fate of Babylon, civilizations that worshipped material gods at grand temples. Working in tech, I value material progress, but those childhood tales remind me: we mustn't promise digital salvation when we're really building better tools.

I personally believe AI is at its best when providing humans with tools and extended capabilities that help them in their work and leisure, it’s not a “thing with true agency", despite moving into the agentic AI world. I suppose over these years I’ve become an AI humanist! (NB, Someone who believes AI's greatest promise lies not in replacing human judgment or creativity, but in amplifying it.)

I do not believe that AI is going to take over the world "Terminator" style, but equally, none of us are naïve enough to think it can't accelerate bad things in the hands of the wrong actors – as an example, a far more rapid ability to generate polymorphic cyber security attacks, or phishing emails with correct spelling that makes them harder to spot.

There are numerous different flavours of AI technologies, but each one comes from a very long evolution of human endeavour. Whether it’s your favourite language model [LLM], the image recogniser on your phone, the recommender system that finds the next video to watch or even a cutting-edge Reinforcement Learning based protein folder model, they are all the work of decades of evolution and human effort. - From theory to algorithm, from GPU development (video game industry) to mass data labelling projects for A.I. training... and that’s without thinking about the comms. revolution that is the internet, which without its ability to share data on a mass scale we certainly wouldn’t have contemporary AI.

It’s all too easy to anthropomorphise your preferred language model. However, to an AI engineer, it is a distributed mass of tensor-based compute operations, concurrently serving millions of other users with its unimaginably rapid token generation capability. Its human “voice” being a product of teams of communication experts fine tuning its raw output to a more human palatable tone – layer upon layer of compute and algorithm to put characters on to your screen in a form you can relate to.

Talented product designers are great at that – they paint happy faces onto vacuum cleaners or children’s suitcases, because it engages the customer.

Despite all those hours studying and working in the tech. industry I genuinely cannot predict how AI will change the next few decades . I also don’t think any tech. leader, publisher of books, expert or guru can either. Social trends, unexpected events and the very nature of our complex planet’s ecosystem means it won't be exactly as we forecast. What we do know is there will be both good and bad. I truly hope that far more people will put AI technologies to good and creative use, than those who use it to generate fake and poor crud. (For which we will need more AI technology to filter out for us.)

As leaders in business and technology, we stand at an inflection point. AI isn't the deity some evangelists proclaim, nor the demon others fear, it's a powerful tool born from decades of human ingenuity. Our role isn't to predict an unknowable future, but to guide AI's integration with wisdom and purpose. Just as the digital revolution transformed business in ways we couldn't foresee, AI will also surprise us.

If we remain grounded in human values, focused on genuine utility over hype, and committed to enhancing rather than replacing human capability, we can help ensure that this ever-evolving technology serves our best aspirations. Not utopia, not Babylon's fate, but positive progress guided by our humanity.