There was a time was when “artificial intelligence” (AI) was an arcane academic subject, beloved by people who enjoyed arguing about the “Turing test” and the Searle’s ‘Chinese Room’.
Now it has burst out of the lecture hall and the university laboratory to become one of the hottest investment trends of modern times.
‘How to’ guides appear in financial research, schooling readers on the mysteries of investing in AI. Top tips come and go as frequently as predictions as to the next AI wonder that is just around the corner, whether that be self-flying passenger planes or affordable domestic robot servants.
For the individual trader or investor, it may all seem a little daunting. It is easy enough to understand what, for example, a company like ExxonMobil does for a living. We may not be able to drill for oil ourselves, but we get the general idea.
AI, by contrast, is unknown territory, developed by specialists and in a constant state of flux. Perhaps you may think that it is best not to invest in artificial intelligence.
The fact is, however, that the same could be said about the pharmaceutical industry, yet layman manage to trade or invest successfully in giants such as GSK. Furthermore, the tried and tested principles of investment have not been suspended by the lightning pace of technological change – they are as relevant to AI investing as to any other industry sector.
More on that in a moment however. First, what exactly is AI? There’s no watertight definition, but we tend to know it when we see it. Conversely, we tend to also know what it isn’t.
Put simply, an electric pump will automate a process previously undertaken by a human, but it’s not AI. Neither is a desktop computer, however prodigious its memory.
War of words
By contrast, Tesla’s new “enhanced summons”, by which its cars will weave themselves through a car park to pick up their owners, very definitely is.
Talking of Tesla takes us to an important distinction in the world of AI shares, between companies, such as the carmaker, that use AI in their wider activities and those that are “pure” AI specialists.
In this latter category is Nvidia, a California-based company that designs graphics processing units for the gaming and professional markets. Interestingly, it also designs chip units for the mobile computing and vehicle markets.
This row may seem comical, but these dizzying speeds are a key component of AI. Without ultra-rapid processing, AI simply would not exist.
In the first category is Amazon, the start-out online bookseller that has become an e-commerce colossus. Amazon is also a leader in AI.
In May this year, Forbes magazine declared: “Their machine already outranks most of the competition in product search inquiries, personalised recommendations, demand forecasting, fraud protection and warehouse fulfilment.”
Its personal assistant, called “Alexa”, which can be ordered to turn lights on and off, uses AI.
On the other side of the Pacific, China’s Tencent describes itself as “an internet-based technology and cultural enterprise” whose mission is to “improve the quality of life through internet value-added services”. One of its slogans – “AI in all” – says it a lot about its commitment to artificial intelligence.
Such is its market position in AI that the Financial Times published a discussion with senior vice-president Seng Yee Lau on 1 May: “In a rare interview, China’s most valuable publicly listed tech company said it favoured an approach to ethics that combined not only socially beneficial uses of AI for medical purposes or agriculture but also ensured a ‘social contract’ between companies and users to govern the use of personal data.”
Talk of ethics leads naturally to potential pitfalls for those investing in AI.
Whether it is the fear of self-driving cars piling into pedestrians or advanced computers being misused by governments to spy on their citizens, many see the whole field of AI as being fraught with investor risk. Huge legal uncertainties remain as to where liability would rest in cases of either accidental or deliberate misuse, and it is entirely possible that the courts would decide that software developers or companies using AI in their products, or both, would be to blame.
A final pitfall may lurk in the very real danger that AI has been overhyped, making the stocks overpriced and vulnerable to a sharp correction.
One principle would be to go for market leaders whose positions are hard for rivals to storm. Microsoft, which has poured resources into AI and developed six principles to govern its use, would be an example here.
Another would be to go for a relatively obscure name that has nevertheless built a big reputation in the business. Advanced Micro Devices would fit this bill.
Finally, however dazzled you may be by developments in artificial intelligence, remember that as a trader or investor it is your own intelligence that matters, and humanity is some way off designing a machine that can match that.
Critics of the Turing test use the idea of a “Chinese room”, one filled with cut-out characters from the Chinese alphabet which an English speaker arranges into coherent sentences having been given instructions in English.
In no sense, they say, is that person “writing in Chinese”.
A fascinating philosophical debate, but of limited use in guiding your investments.