In downtown Tokyo, just like in most major cities, taxi drivers make an educated guess as to where they might find their next paying passenger. Years of experience have honed their prediction skills to pinpoint customers depending on location, the time of day and weather. But even with this knowledge, tracking down passengers is still a hit and miss science with many drivers saying they can cruise around for up to two hours without finding a fare.
Japan's biggest mobile phone operator, NTT Docomo, wants to change all that. The phone giant says it is in the process of creating a transportation revolution by perfecting a new taxi passenger pickup system using Artificial Intelligence (AI). This system predicts future taxi demand and directs drivers to high potential locations using AI.
It does this by splitting the taxi driver's chosen area of operation into a 500-meter grid pattern that is displayed on a dashboard fitted iPad-sized tablet. This shows the number of customers that may use the system in the next 30 minutes in a particular grid.
Incorporating passenger pickup and drop-off data from 4,425 taxies, in addition to time of day and weather factors, NTT Docomo trained its AI technology over an 18-month period to predict where potential passenger concentration will be most active within the next 30 minutes. Obviously NTT's greatest strength is that it can use its WiFi network to show the location of customers using their mobile phones.
One taxi driver I spoke to said he had noticed a 20% rise in passenger traffic since using the AI prediction system. The more a taxi drives around, the more the AI system learns by itself, giving it the ability to predict where a customer will be with an accuracy rate of up to 95%.
Now that's all well and good. But hang on. It learns by itself? Shouldn't that be raising alarm bells? While some privacy proponents have voiced concern about the sharing of personal data and potential cases of fraud, others are worried about the extent that AI can self-learn -- and act.
The first place AI prediction software appeared was inside Nomura Securities, Japan's largest financial services group. Every 1/1000 of a second, the company's latest AI software measures hundreds of variables in deciding how much a share price will fluctuate and then it suggests a plan of action every five minutes. "It arrives at analytical solutions that humans were not able to even contemplate up until recently. We can no longer do without AI in this line of business," says Taishi Harada, Nomura's AI Implementation Director.
So AI can arrive at solutions that humans were not able to reach? The AI story gets even cloudier when we examine disturbing news out of Japan recently that Ponanza, an AI-wielding Shogi (Japanese chess) robot beat a professional master player at his own game. The fact that the robot defeated the master was not the story. The scary comment by the robot's creator, Issei Yamamoto, that he could not explain how the robot arrived at its game-winning strategies sent shockwaves through Japan's media networks.
It reminded me of the telling lines in the 1984 flick The Terminator when Michael Biehn's character Kyle Reese explained how the apocalypse started. "Defense network computers. New... powerful... hooked into everything, trusted to run it all. They say it got smart, a new order of intelligence. Then it saw all people as a threat. Decided our fate in a microsecond." Sound like an early definition of AI? Um, yeah!
Now we have not reached that point yet with AI. But Tesla boss Elon Musk and Professor Stephen Hawking are concerned. They have both declared AI to be the most serious threat to the survival of the human race. Musk says "humanity risks summoning the demon with AI," and has stressed its development must proceed with a strict set of checks and balances, something that is not happening now.
Both visionaries agree that while AI technology can be harnessed to help humans in areas such as medicine and care for the aged, AI's ability to self-learn and act on that new knowledge is a real and present threat and needs immediate and tough governmental regulatory oversight.
Meanwhile, back on Tokyo's roads, Skynet is not about to take over. If AI can innocently assist passengers to find a taxi ride easier while boosting cab drivers' take home pay, then it's got to be a mutually beneficial relationship. Right?