AI - not the Fed - will bring down inflation
Better quality and lower prices are coming in Health and Education.
The professional classes are currently, and rightly, obsessing about the impact that AI will have on us as Chat GPT get exponentially smarter. Investors meanwhile are trying to work out who the ‘winners’ are going to be and are chasing tech stocks in the US higher. I suspect the impacts will be profound, but probably not in the way most of them are thinking. In the first instance I would see it as significantly improving living standards for the 99% at the expense of the 1%, in particular by collapsing service sector inflation and making many services both more efficient and more effective.
The current iteration of Chat GPT, the most well known AI, is said to have an IQ equivalent of around 155, meaning that the next one will, quite literally on this measure, be the ‘cleverest’ thing in history. Much is currently being written – a lot of it almost certainly by Chat GPT itself – about how it will put various white collar employees out of a job, but in fact I suspect that its first impact will be to dramatically improve the productivity of many in white collar jobs and that rather than being made redundant, it will instead mean that the supply of services will increase and relative wages decline.
‘Man plus machine’ means that the first impact of AI will be to dramatically improve white collar productivity. This means better products at lower prices
It has always been this way, since the industrial revolution, bottlenecks in production and spikes in the pricing power of labour have resulted in innovation and brought in capital that has led to more and better products at lower prices. Thus in the 17th century when the invention of the spinning jenny led to an excess supply of yarn compared to the number of weavers, the weavers apparently took to wearing top hats with five pound notes in the hat band. That is until the power loom changed their pricing power. This has been a continuous thread in economics and of course in markets which provide the capital to bring about this innovation and in the latest episode of globalisation it has actually been automation and robotics rather than cheap labour that has been a major driver of ‘stuff’ being both cheaper and higher quality than it has ever been. A lot of this automation has come from ‘tech’ of course and we have already seen this pricing power effect at work in many areas of the service sector thanks to digitisation; books, movies and music have all been massively disrupted to the benefit of consumers versus producers – including, importantly, a lot of the ‘middlemen’ - and new distribution channels have emerged.
As has been noted, the real winners from digitalisation did not even exist as the dot com bubble burst and it will be the same here. We believe that AI will similarly change the pricing power in other, key, parts of the service sector and thus that one of the first, and so far not really discussed, impacts of AI will be to lower inflation rates in developed countries and dramatically increase the living standards of the 99% of the population who consume those services, mostly at the expense of the 1% who provide them.
AI will dramatically improve the lives of the 99%, by lowering prices and increasing quality of key services like health and education
In less developed countries it will enable the whole population to rapidly close the gap with developed countries in many areas, particularly where the high cost of services is limiting access. So if AI is about to bring this level of disruption to the service sector, bringing higher quality, more competition and lower prices, where is it going to have the biggest impact? To me, the most obvious place to look is in areas that have so far largely resisted the deflationary impact of digitalisation and where the pricing power of labour appears to be the highest.
Thus, if we look at the CPI in the US over the last two decades, it should be no surprise to see that the biggest drivers by far were health and education, where practitioners have managed to retain exceptional pricing power, through amongst other things a combination of unions and restricted labour practices, drug patents, technological barriers to entry, regulatory monopolies, lobbying and general ‘guild’ like behaviour. It is these areas that are about to be turned upside down by AI.
Digitalisation produced the blue lines - AI will transform the red ones
During Covid, the doctors and the teachers told us that we could do almost everything online, true, but they had, perhaps unwittingly, just undermined the very foundations of their guilds. If we can watch lessons/lectures online, then why wouldn’t everyone want to watch the very best lecturer/teacher rather than be restricted to watching the bored (and boring) tenured professor recycle notes from the previous year(s)? Incidentally, why, apart from guild behaviour, do we still produce and consume school textbooks? The AI will be the smartest teacher ever known, so the Phd student who currently runs the tutorials, or the teacher’s assistant who currently helps with homework, when armed with the AI completely reverses the power structure. The pricing power now will be with the empath who works closely with the student/pupil, not the ‘academic high priest’.
Just as tech empowered the ‘nerd’ over the ‘jock’, so AI will empower the ‘Empath’ over the nerd.
Similarly, a large part of the diagnosis side of medicine is pattern recognition, something that computers excel in but AI will undoubtedly be superior to humans at doing on account of having billions of data points constantly updating and ‘learning’ as it goes. Another reason for the high cost of medicine is the increasing use of diagnostics equipment that is not only expensive, but under- utilised due to the bottlenecks in processing the outputs. AI can vastly improve the efficiency of the medical specialists involved as well as increasing the pricing power of empaths over ‘boffins’. This will enable the consumers of medical services to have quick access to higher quality care and for competition to bring down prices. Efficient and effective. Of course, AI is not the only technology that is going to do this, ‘wearables’, smart devices – including smart toilets (!) will be able to monitor our ‘vital signs’ constantly, while the AI can constantly monitor for patterns and diagnose solutions. Early diagnosis and prevention rather than hospitalisation or expensive drugs that , as the saying goes “work in one third of patients, have no effect on the next third and are actually harmful to the final third’ will dramatically shift the cost benefit analysis in medicine. Meanwhile, the power of AI will also allow for a massive increase in the development as well as the efficacy of new, more tailored drugs.
Right now investors are betting on facts like AI will use a lot of processing power and a lot of energy, but it is pricing power rather than processing power that we need to think about.