During the Warring States period, the King of Wei one day had a whim and gifted Huizi a large gourd seed, which sparked a legendary debate between Zhuangzi and Huizi. Huizi was dismissive of the king's gift, saying that the large gourd grown from this seed, if used as a water container, would be too fragile despite its size; and if made into a large ladle, there would be no correspondingly large vat to hold it — making it a truly oversized and useless thing. This argument drew mockery from Zhuangzi, who said: "Huizi! You simply do not know how to make use of large things."
Whether in life or in the workplace, it is not uncommon to encounter situations where the same thing looks completely different depending on one's perspective. I once met a senior executive responsible for big data, who told me that his current data analytics team had no need for machine learning or for learning the new programming language Python. What he meant by "no need" was not that he had some different idea about choosing better AI tools — rather, he believed his team did not need to learn AI to change the way his existing team worked or to improve their processing capabilities and efficiency. In his view, his existing manual, rules-based approach using tools like SQL and SAS was sufficient. AI and machine learning were, in his eyes, like Huizi's oversized gourd ladle — big, but utterly useless.
In my view, this mindset is not only a shame — it is alarming. It is a shame because, at a time when the data economy is just beginning to take off, this company's data analytics team has already stopped growing, simply because its manager does not know how to leverage AI and machine learning and has cut off his team's opportunity to learn how to mine the gold buried in data.
It is alarming because Taiwan may have lost yet another company capable of giving Taiwan's young talent a reason to stay. If this manager were willing to understand what today's young people — those striving to become data scientists and devoting themselves to the data economy — are working so hard to learn in terms of new data science techniques, he would surely make a different decision.
Middle-Aged Managers: Are You Just Waiting to Retire Comfortably?
In Taiwan, we often say that young people are obsessed with small comforts, are dependent on their parents, immature, angry at the world, raised in an overly relaxed environment, and lacking drive. But I have also met outstanding Taiwanese students who, despite having backgrounds in humanities or business, spend three hours every day teaching themselves Python and machine learning. If anything, I believe the ones who most need to pursue learning and growth, shift their mindset, and stride boldly forward are people of my own generation — the middle-aged managers. Are these managers, with their conservative attitudes, slowing down the pace of progress, trying to preserve a comfort zone, and simply waiting to retire comfortably in the next five to ten years?
Returning to the middle-aged manager's core business — retail banking — it has four defining characteristics: 1. a large customer base; 2. a wide variety of products and services; 3. many friction points; and 4. an abundance of data. Advanced retail banking technology is inevitably a domain where AI and big data can generate tremendous positive impact.
Why Must FinTech Learn AI?
Many of these middle-aged managers may have forgotten that since the year 2000, online banks have been eagerly pursuing "one-on-one service through digital channels." Now, precisely this kind of service is finally seeing the light of day. Thanks to the diverse integration of banks with other industries, the variety and opportunity of leveraging cross-industry data have expanded enormously. The retail banking customer value chain — spanning customer needs exploration, marketing, customer acquisition, product and profitability management, and in-process and post-process service management — has unfolded like a peacock spreading its tail, branching into a decision tree with thousands of unique faces. With tens of thousands of decision points, building these using traditional manual modeling has become far too slow to keep up.
I once pitted manual modeling against machine learning head-to-head and found that when the number of variables increased from 10 to 30, the complexity did not triple — it increased thirtyfold. In the world of big data, we will be dealing with 3,000 or even 30,000 variables. The diverse combinations of algorithms make it entirely impractical to rely on traditional manual predictive modeling. Empirical evidence has shown that machine learning versus human expert modeling results in machine learning saving at least two-thirds of the time while achieving the same levels of accuracy and stability.
In the process of internal corporate innovation, innovation teams will inevitably make mistakes. When accused of pursuing something oversized and impractical, all innovation teams should humbly take the criticism on board. But the greatest challenge arises when an organization becomes internally divided into two worlds — teams with and without innovation responsibilities — potentially cultivating within the organization another group of Huizis, passively waiting to be convinced: "What use is AI and machine learning?"
As for young Taiwanese people whose managers resemble Huizi rather than Zhuangzi, do not be discouraged — the world is on your side. On the Coursera platform, Stanford University offers a free Machine Learning course that has attracted more than 2.2 million students from around the world. If you search for "Machine Learning" on Coursera, you will find more than 200 courses offered by world-renowned universities. If you want hands-on practice, there is also Kaggle (already acquired by Google), a platform that hosts a wide variety of prediction competitions and datasets using AI and machine learning across fields including health, medicine, finance, insurance, and consumer industries, with prize money for individual competitions reaching hundreds of thousands of US dollars.
Be a broad-minded middle-aged manager like Zhuangzi; be a young person who knows how to make good use of external resources — and the world will be ours.
Source: https://www.cw.com.tw/article/article.action?id=5094915