Fintoism Blockchain AI Congestion

AI Underperforms Aganst Human Analysts In Money Making In Financial Markets

Artificial intelligence is an exciting technology with tremendous future potential. Several financial institutions leverage AI to make money or help clients do so. Surprisingly, humans still perform far better than AI solutions on this front, although that situation may change in the future. 

Humans Beat AI In Money Making

It is remarkable to see humans stay ahead of artificial intelligence when it comes to making money in financial markets. Those are the findings of a new research paper. Moreover, the paper questions the machine learning algorithms deployed to make money in financial markets and whether sufficient research has been done in that field. 

With so many financial institutions embracing AI for chatbots, cybersecurity, advisory services, and many other use cases, one would expect artificial intelligence to perform better. However, the study shows there have been a fair bit of cherry-picking results in previous studies. Moreover, running multiple versions of an investment model in parallel creates a bias only to present the best-performing outcome and ignore the rest. 

Such a parallel approach would not work in a real-world investment management setting. You can only execute trading strategies once, and they have to be up to par to record a profit. One cannot undo results after the fact, and they are unlikely to be as positive as several research studies would like the world to believe. Moreover, there is a lack of acknowledging AI algorithms have a black box nature and don’t always consider regulatory restrictions. 

The data taken from funds publicly disclosing market data makes things even more interesting. All of those funds underperform compared to humans. That further indicates human analysts will have job security for some time to come in this field. However, ongoing improvements in the AI field can turn this situation on its head rather quickly.