We had intended to retire the AI but following some quite positive feedback (which we don’t usually get) it gets a reprieve. We also get the sense that readers might be as interested in our exports in the land of AI as our pontifications about markets, which we wouldn’t take badly at all. Makes sense, as we all have to work out how to use this stuff and we don’t mind being your guinea pig if you don’t mind returning the favour. A few things stand out:

• The program that we have ended up using most for this purpose, which in turn uses the ChatGPT4 engine, really does write quite well. It’s also succinct (we’d elaborate but it turns out people don’t like waffle. Who knew.)

• The more you use these things the more the magic does start to look a bit more like plagiarism. It works best and is most reliable when you prioritise articles that it finds for you and also add some off your own (or others’) material. So, props to T. Rowe Price in the US for being what we think was the biggest influence on this week’s AI generated summary. We also fed it a podcast from Milford Asset Management (whose content was also quite obvious – we hope they don’t mind).  

• And that’s where it gets spooky again. Some comments were obviously, and sometimes rather clumsily, paraphrased (as one often sees in teenage homework when big words suspiciously take the place of simpler ones). But in other areas, it seems to draw some surprisingly logical and common-sense inferences that are not obvious in the text that has been focused on, implying that a more widely formed view is being brought to bear.

• Lastly, we think it does a good job of dynamically organising the data into structure that reflects the content and then focusing on the dominant themes. Last week, for instance, it organised the data by region as there were some markedly different regions effects but this week it identified that a more representative structure would show a dichotomy between the common voice emerging from central banks and the conflicting messages coming from the real economy (equally common across different regions).          

You can read the AI Summary here, but the overall narrative was that central banks continue to fight inflation and worry about obviously tight real time labour markets while leading indicators point to a slowing economy and looming recession. In that light it makes sense that most equity markets were on the back foot, while bond markets were surprisingly resilient to the prospect of higher rates – ten-year rates were down around the world while short-term rates rose slightly. Commodity markets also marched to the recessionary drumbeat with most of the energy and metal complexes down a few percent, although Iron Ore has recovered in the last few days. The biggest equity falls (around 4%) were from emerging markets, led by China. Japan also fell by 3% and Europe (including the UK) was down by 2% while US markets were only down 1% (with tech stocks and industrials trading in line for change). The week before last markets had been buoyed by the prospect of Chinese stimulus, but a lack of follow-up from the authorities has dampened share market expectations and possibly the prospects of its largest export focused trading partners which could add to the bite of domestic recessions in the West. Lastly, credit markets at last started to sniff out a recession, with yields easing last week but only by a few basis points for now.

‍