Season 6, Post 47: The silver bullet

Everyone has strong opinions about AI, yet we all still want to learn more about this potentially transformative technology. Perhaps it’s because it addresses itself to “everything that requires intelligence in human beings,” in the words of Michael Wooldridge. His is a voice worth listening to, for he is the Ashall Professor of the Foundations of Artificial Intelligence in the Department of Computer Science at the University of Oxford. He has advised the UK Government, among other parties, on AI. Your author was privileged to attend a talk from Michael last week and found it highly illuminating.

AI constitutes “a watershed moment in tech history”, according to Professor Wooldridge. He then paused. The audience knew that there was a ‘but’ coming. “But there are some issues”, he added, with the distinct suggestion that this may be an understatement. Two specific issues that he called out at different points in his talk were that big tech “sees gen-AI as a silver bullet” while governments “hope that AI will be a silver bullet.”

As an exercise, your author asked ChatGPT to explain the concept of the silver bullet. My handy LLM detailed that the term typically “refers to a simple, seemingly magical solution to a complex problem”, implying a one-size-fits-all outcome. Sadly, silver bullets rarely exist for real-world problems just as outside the realms of folklore, werewolves, vampires and witches (whom, it was said, could only be neutered by such a bullet) do not exist either. 

We may then be placing excessive hopes in a novel and largely unproven technology. Our own near-term AI scepticism has been voiced for some time (and elaborated on in our major thought piece in this year’s theme book). For Professor Wooldridge, “nobody knows what’s going to be the killer app”, but by big tech seeking to “embed it everywhere”, the hope is that they will be able to find it. Hope, of course, is not a strategy. Equally, a country’s productivity problems won’t, in our view, be solved by a single technology alone, particularly when its implementation may take time and will almost certainly have unintended consequences.

One of the more intriguing ideas developed by Professor Wooldridge was the notion that within the next decade or so, we conceivably will not know the difference between human and AI-generated content. Regular readers may be familiar with your author’s written style, but ChatGPT’s responses in our dialogue about silver bullets above appear as unerringly human. Look forward and LLMs will only likely get more powerful. Imagine their potential once they can capably manage to interpret unstructured data, power robots and make virtual reality experiences more compelling, to name but a few examples. Do not forget, however, that AI will still “learn to do things that we cannot always predict.”  

3 December 2024

The above does not constitute investment advice and is the sole opinion of the author at the time of publication. Past performance is no guide to future performance and the value of investments and income from them can fall as well as rise.

Click to here view all Blog posts.

Alex Gunz, Fund Manager