Imagine you’re planning on investing your money on the stock market. Before purchasing any stock, you want to gather enough information about which stock is reliable and worth your money. This personal touch depends on the information available and your interpretation of it. To create an opinion time is needed. But is this information up-to-date? Do you interpret the data correctly? Now Imagine someone who does it for you and makes this decision in a second instead of 2-3 days. Wouldn’t that be great?
Roughly a decade ago the stock market started losing its human touch by the introduction of algorithms run by computers. Using algorithms, tremendous amounts of calculations could be computed which predicted future stock behaviour within little time. A term flash-trading was introduced which referred to these programs buying/selling stocks every second.
In recent years the development of artificial intelligence has been branded as the next investing oracle. Goldman Sachs alone had 831 job listings in search of programmers or big data specialists[1]. Artificial Intelligence is able to gather and analyse more data and it applies machine learning. Currently owning an infrastructure capable for running artificial intelligence and machine learning requires a large investment. But as the costs of hardware keep falling, artificial intelligence will be easily accessible for all[2].
But how do we stay ahead if artificial intelligence programs become available for all? By fully relying on these prediction machines, the stock market is at risk of losing competitive advantages and innovation. Even one of the most innovative men, Elon Musk, claimed robot involvement within Tesla had gotten out of hand and humans had to be brought back[3].
A perfect explanation of machine imperfection is given by Joshua Gans a professor at the university of Toronto[4]. US air-force planes were often shot down during bombing flights. To prevent planes from being shot down, the planes could be reinforced. However reinforcing every part would make the plane too heavy. Would the solution be to reinforce area’s which were hit most? Abraham Wild, an famous statistician, told the US Air force the opposite. Reinforce the parts which are least damaged. These probably are area’s hit of planes who do not return. Machines cannot react on data they never received from the planes which were shot down. Such a decision would be highly counterintuitive for a computer.
To conclude, the possibilities of stock market prediction using artificial intelligence are enormous and extremely beneficial. No human could ever process that amount of data in such a short time. However, the risk remains a machine can miss critical data which seems random or which it does not receive. A humans touch on algorithms run by machine still seems to be the competitive advantage in the stock market industry.
Would you fully rely on artificial intelligence to run? Or do you believe human touch is key to gain a competitive advantage?
Sources:
[1],[3],[4]https://www.marketwatch.com/story/ai-will-change-stock-market-trading-but-it-cant-wipe-out-the-role-of-people- 2018-05-15
[2]https://www.itprotoday.com/machine-learning/how-ai-trading-systems-will-shake-wall-street
I think the question you pose regarding the influence of artificial intelligence on the stock market is interesting, as 54% of Americans are invested into the stock market through individual stocks, mutual funds, pensions or retirement plans (1).
Artificial Intelligence has significantly disrupted the stock markets already with the introduction of index funds, which is a portfolio constructed to match or track a market index, providing broad market exposure, low operating expenses and low portfolio turnover (2). These index funds are often fully operated by machines and rely heavily on Artificial Intelligence to track a market index. These machines grew exponentially in the last decade and are now responsible for well over 4 trillion USD (3).
Thus it seems that humans are okay with fully relying on Artificial Intelligence, even when it comes to their money! It will be interesting to see which industries will also be disrupted by Artificial Intelligence!
Sources:
(1) https://money.cnn.com/2017/10/20/investing/trump-stock-market-americans/index.html
(2) https://www.investopedia.com/terms/i/indexfund.asp
(3) https://www.ft.com/content/89c18106-3591-11e7-bce4-9023f8c0fd2e
You have touched upon a very interesting topic in my opinion Maurits thanks. The continuous improvement of AI makes it a big challange for white collar workers in terms of efficiency for sure. And in the case of investing your money, they also have the advantage of being unbiased compared to a human. Trusting a machine with your money can be a better option than trusting someone you only talked on the phone for seconds, considering that the machine runs countless operations and algorithms simultaneously to provide you the best alternatives while the human relies mostly on their persuasive skills. However, this strength of AI also brings a major downside with it. Since it runs on the data it is given by various entities (individuals, corporations etc.) within the system, a security concern may rise. In case of a wrong or manipulated data entry by any of the entities within the system, results of analysis done by AI can be manipulated to a great extent. Additionaly, considering that the power of rumours and speculations in the stock market are still major factors of making right and wrong decisions and humans still possess the advantage of exploiting these powers. As a result, it again comes down to enabling AI and humans work together for a common goal, with each of these factors using their own competitive skills; for example in the case of stock market a good scenario would be where AI runs simultaous analysis while humans deal with the communication with customers. My first blog entry deals with this issue in detail. I am copying the link below, hope you can provide more insights on the issue based on your research 🙂
https://digitalstrategy.rsm.nl//2018/09/06/the-future-of-the-human-factor-in-business-with-ai/
Thanks for you insightful reply Hayrettin. I did not yet think of the possibility of the dispersion of false data and its effect on the entity using AI. Very interesting! Il give your blog a read!