Go, an ancient board game popular in China, Korea and Japan, is far more complicated and unpredictable than chess.
Which is why AlphaGo’s four-to-one win against a human Go master marked a giant step in the advance of artificial intelligence.
“There are only about 20 possibilities on average for a move in the case of chess, but for Go, there are about 200,” Chordio Chan, head of investment at Bank of China (Hong Kong), writes in his Hong Kong Economic Journal column.
The goal of chess is more defined, which is to capture the king, but Go players place stones on the intersections of a 19 by 19 grid to compete for claiming the most territory.
“It’s much more difficult to quantify the decision process,” Chan notes.
“What’s so amazing about AlphaGo is not its brute-force computing speed and power, but the refinement of its computing method.
“It is capable of learning from matches and adjusting its strategy.”
AlphaGo’s programming is said to be modeled on neural processes to replicate human instincts.
Perhaps not as sophisticated as AlphaGo, but artificial intelligence is already playing a big part in the financial world.
Chan cites the case of American financial data services firm Kensho, which has developed an analytical tool targeted at institutional clients.
Kensho calls it “Warren”, coincidentally the same as billionaire investor Warren Buffett.
Warren can quickly answer 100 million questions that may impact share prices. It is also capable of writing detailed research reports, and it can raise the efficiency of back-testing investment models with historical data.
Kensho already secured investments from Goldman Sachs, Google and CNBC.
Artificial intelligence may bring sweeping changes to the investment banking industry in two ways, according to Chan. One is the use of the technology in big data analysis, or robotic analysts. The other is the adoption of artificial intelligence in money management.
Robotic fund managers can sharply reduce the cost and hence the fees to under 0.5 percent. Minimum investment amount can be as low as US$10.
Perhaps the best thing about robotic fund managers is their emotionless approach. As seasoned investors point out, it’s often an investor’s character that determines the outcome.
“Be fearful when others are greedy and greedy when others are fearful.”
This famous quote from Buffett could be the secret to profit, but how many can put his words into practice? This is just against human nature.
Robotic investment consultants are expected to manage US$300 billion of funds by the end of this year, according to consultancy A.T. Kearney.
Rapid growth will take the number to US$2.2 trillion by 2020.
Charles Schwab is one provider of such service, Chan notes. Based on answers to a questionnaire covering items like investment target, size of fund and risk tolerance, the company’s system automatically generates a recommended portfolio.
Charles Schwab’s robotic manager will also keep track of the performance and make ongoing adjustments.
A Skynet-like artificial intelligence system that aims to destroy humans, as featured in the Hollywood hit movie The Terminator, may not happen, but intelligent robots are visibly challenging humans in a wide array of fields and jobs.
As computers get smarter, humans will be hard-pressed to identify areas where we can still excel.
– Contact us at english@hkej.com
CG