I was late to the end of the start of AI*

*This is not entirely true but it's a snappy title.

Artificial Intelligence changes the entire game

I don't say this lightly; to put such a statement in context and give it it's due import, anyone who knows me knows that I have a finely tuned BS detector. This is partly because I come from a culture that has its origins in a penal colony and partly because I'm becoming more curmudgeonly in my advancing age. I'm a reformed cynic but my scepticism is borne out of working for multiple IT vendors. This taught a young and naive technologist an important aspect of selling (i.e., don't confuse selling with implementing).

This time you should believe the hype

Please allow me to explain why.

AI happens while you're busy doing other things

In recent years, and up until early 2025, I used to joke that "AI is so profound that I don't cover it". What I meant by that is I couldn't do it justice. In reality, it masked a deeper truth that I was too lazy to keep up given everything else I had going on in my life.

I was right there with Roger Murtaugh (played by Danny Glover) and his oft-repeated quotes from Lethal Weapon: "I'm too old for this sh!t" and having just "eight days to retirement" [1].

This is your wake up call

I'll be honest, nothing shocked me out of my comfort zone more than Google's NotebookLM. What blew me away was the ability to upload and analyse content via a simple interface requiring zero knowledge of AI. Prior to this I'd had numerous small eureka moments (e.g., a colleague saying "just get an AI to write the PowerShell script for you" and an Australian bank explaining how the Vietnamese financial regulator needed them to reinterpret all local language contracts whenever there was a change in legislation or regulation).

Some time soon after my NotebookLM moment, I realised that I couldn't rely on my career's experience to deduce what was happening. For the first time in my professional life, I didn't know how the technology was doing what it did. Sure, I'd studied neural networks and linear algebra (more on that in a moment)...

...but it was, effectively, magic.

As a scientist, this made me feel very uncomfortable.

Back to school

With time to spare — for the first time in a very long time — I "hit the books". I took online courses that taught the nitty gritty of transformer architecture, named entity recognition, parameter efficient fine tuning, self attention, tokenisation, quantisation etc etc.

And I'm just getting started.

Machine-learning models are just big statistical calculators that work with numbers rather than words - Not sure ^.

^ My notes suggest it was someone in this course (but I can't find it otherwise I would attribute appropriately). In all cases, I like the statement (noting it is slightly modified for humorous purposes).

If you want to know how LLMs work "under the hood", there's a lot to learn. In point of fact it took me more than 30 years to realise why my university forced me to study linear algebra (e.g., matrices, vectors). These studies helped me to deduce that current generations of AI were built on information science and statistics / probability but that was just an educated guess that might have been wrong.

In some respects, I've been in the AI field since studying it at university in the early '90s. In fact, I was getting paid to use Prolog at work just a few years later.

I recall learning how Japan declared the 1990s as being the "decade of AI" whilst I struggled to keep up with my team as we battled against other students (to win a game of reversi). Ironically, the winning team trained their program to beat ours; we came second (no thanks to me).

It's deja vu all over again

(Sic)

Fast forward to one of my stints as an IT industry analyst. I was presenting on the top trends facing Asia Pacific organisations in 2014 and beyond. I presented the same talk in Hong Kong, Tokyo and Beijing (among other cities IIRC). Among the trends that I mentioned were the labour impacts from digitilisation and the rise of machines. In the aftermath of movements such as Occupy Wall Street, many of trends touched a nerve.

Among all the content — now over a decade old — a few things stand out. Most notably, the experience of having a piece of technology beat you in your field of expertise.

This [is] what it looks like when the future comes for you ... it’s not a Terminator laser eyepiece tracking you down in an alley; it’s just a line on a chart gradually but inexorably moving closer and closer to human performance — to the thing that you can do that you think makes you special. - Ken Jennings (Source: Bryant University News)

Ken has already experienced what most of us can't yet imagine.

Stay tuned. We're just getting started.

Notes:

[1] All credit to Adam Spencer who referenced this same phrase, in this same context, in a presentation on AI in Sydney in May 2025.