The Taiwan Banker

Two years after Waymo started its self-driving taxi service in San Francisco, California, thousands of Apollo Go self-driving taxis began serving Wuhan, China. These two services were both strongly protested by human drivers, demonstrating that AI has seized a new type of work. With the rapid development of AI, the future impacts on society will challenge our imagination. These self-driving taxis represent a breakthrough in trust in algorithms.After all, they not only need to gain the trust of their passengers, but also other drivers and pedestrians. In the commercialization process, they must prove that they are as safe as human drivers, thereby promoting the transformation of the urban model. The financial industry has also followed these developments. After analyzing collision data, the leading insurance Swiss Re entered into agreements for self-driving taxis with Waymo and Baidu (the parent company of Apollo Go). This data helped Waymo in its operation application process. The transformation in trust from humans to machines eventually resulted in systematic trust in AI, independent of human supervision. By extension, the moment is just around the corner when humans will move from trusting medical equipment to relying on AI for diagnosis, prescriptions, and even surgery. This transformation will also have a profound impact on the financial industry. For example, the success of ETFs and robo-advisors has led many investors to trust algorithms rather than professional advisors. In the future, as the digital native generation becomes the main force in the economy, personal savings, credit transactions, and back-office audit functions will all be re-evaluated to determine whether they still have value and whether real people still need to do the work. However, self-driving taxis also show the limitations of AI, especially in dealing with human problems.For example, some autonomous vehicles only drive according to the algorithm, and do not give way to each other, resulting in deadlocks at intersections. Human drivers who fail to follow the rules drag other vehicles into accidents. This all means that AI still has difficulty mastering the ethics of courtesy when driving, let alone recognize dangerous driving by observing words and expressions. This corresponds to the lack of empathy in AI in financial services. Excellent tellers can distinguish which customers are remitting money due to romance scams. This grasp of human nature will be the capstone required for trust in AI. After all, the core of service lies in insight into human nature. Over-reliance on algorithms and machines may cause us to lose sight of the original intention to enrich customers’ lives.No matter how fascinating the virtual world is, it cannot replace the warmth of real interaction. In my personal experience with Waymo, the smoother throttling made corners smooth and comfortable, allowing me to fall asleep while riding. However, in retrospect, driver was just a program with no understanding of driving ethics. I couldn’t chat with it about everyday things, let alone ask it about San Francisco. In the eyes of the program, meanwhile, I might as well not have been a tourist, but just a bag needing delivery to a destination.