During my student days in the late 1980s, I was one of the few students who had access to an Apple II computer. A big difference between Apple’s word processor then and the average word processor on a PC was WYSIWYG, acronym for What You See Is What You Get. On the screen, the document was visible just as it was printed. Even then, Apple was way ahead of the competition.
Now my best investment ever is a course in ten-finger blind typing. At around three hundred strokes per minute, assignments were swiftly completed. Recall that in those days there were still typing rooms where written or spoken texts were worked out. For each assignment, each student had to first enter the library for extensive research on sources.
If a book was not available in the university library, you could always go to the Royal Library in The Hague, where eventually the information you were looking for could be printed off a microfiche after a long search. Thus, for two pages of source material, you could easily spend a whole morning. The big difference with today is mainly in the speed and accuracy of searching. Search engines on the internet work just as fast in their own stored documents in the cloud as on the internet. A morning’s search has been reduced to a few seconds.
New revolution
Now there is a new revolution that is fast becoming popular. It is ChatGPT. GPT stands for Generative Pre-Trained Transformer. It is a chatbot launched by the company OpenAI in November last year. This is a self-learning programme that can formulate good and reliable answers to complex questions. Furthermore, it can even answer follow-up questions and also admit mistakes. With this, the computer is starting to look more and more like a human and, combined with improvements in speech technology, many human tasks can soon be taken over.
Checking an account, making payments and tracking investments is now possible through simple and intuitive questions. But ChatGPT is more than glorified customer service. It can also be used to analyse financial data and is rapidly getting better at it than the average analyst. That’s because one person cannot possibly collect, read and analyse all the relevant information. The machine always has an edge. However, humans can make better decisions based on all that information and therefore work much faster.
ChatGPT is also much better than humans at recognising trends and patterns. Useful for investors, but also, for instance, for all those employees at banks investigating money laundering, for instance, or screening companies based on all the components involved in a sustainability analysis. ChatGPT does have its limitations. It cannot properly distinguish between fact and fiction.
It can be utter nonsense
The output can look good, but at the same time it can be utter nonsense. Furthermore, it is impossible for ChatGPT to depict the future, in which humans differ essentially from machines. Communication, after all, is much more than merely language.
Nevertheless, ChatGPT will ensure that tasks in many service professions can be successfully completed much faster. Incidentally, ChatGPT is not alone. The difference it makes is mainly in its successful output. Earlier, Meta came up with Galactica, an artificially intelligent machine trained by studying 48 million scientific publications, only this machine turned out to spout so much relative nonsense that it was turned off only after two days. Furthermore, Google cannot be left behind and will have to come up with its own version of ChatGPT.
For online investors, ChatGPT could soon provide answers to complex personal questions on pensions, finance and investments. ChatGPT is especially excellent in such a complex environment with many rules. Everyone, including every lawyer is supposed to know the law, but in practice, only ChatGPT is omniscient. Still, the chatbot cannot replace personal financial advice, but it can provide excellent support.
More efficient markets
ChatGPT will help make financial markets more efficient. After all, efficient market theory assumes that decisions are made based on all known information. That has become a lot easier with ChatGPT’s support. Ultimately, just like the Apple II and the advent of the search engine, it is a huge boost to productivity and therefore creates more economic growth. For the record, this text was not written with the help of ChatGPT. Should that be the case in the future, then, if you ask me, it will come with a special disclaimer.
Han Dieperink is chief investment strategist at Auréus Asset Management. Earlier in his career, he was chief investment officer at Rabobank and Schretlen & Co. His contributions on Investment Officer appear once a week.