The Cognitive Limits of Lifelong Learning

关于终身学习的认知局限

Edoardo Campanella 爱德华多·坎帕内拉

As new technologies continue to upend industries and take over tasks once performed by humans， workers worldwide fear for their futures. But what will really prevent humans from competing effectively in the labor market is not the robots themselves， but rather our own minds， with all their psychological biases and cognitive limitations.

In today's fast-changing labor market， the most in-demand occupations– such as data scientists， app developers， or cloud computing specialists – did not even exist five or ten years ago. It is estimated that 65% of children entering primary school today will end up in jobs that do not yet exist.

Succeeding in such a labor market requires workers to be agile lifelong learners， comfortable with continuous adaptation and willing to move across industries. If one profession becomes obsolete – a change that can happen virtually overnight – workers need to be able to shift nimbly into another.

Lifelong learning is supposed to provide the intellectual flexibility and professional adaptability needed to seize opportunities in new and dynamic sectors as they emerge，as well as the resilience to handle shocks in declining industries. Training centers， the logic goes， simply need to identify the competencies that companies will look for in the future and design courses accordingly.

Yet， in the eurozone， only about 10% of the labor force undertook some type of formal or informal training in 2017， and the share declined sharply with age. If lifelong learning is the key to competing in the labor market，why are people so reluctant to pursue it？

Lifelong learning is viewed as extremely costly in terms of time， money， and effort， and the returns are regarded as highly uncertain， especially amid technological disruption. Such views may be reinforced by the feelings of depression and hopelessness that often arise when workers lose their jobs or face career crossroads.

Human beings experience a decline in cognitive performance relatively early in life， with fluid intellectual abilities – associated with working memory， abstract reasoning， and the processing of novel knowledge – beginning to decline around age 20. After middle age， these abilities deteriorate substantially， making the acquisition of new skills increasingly challenging. Only our crystalized cognitive abilities， related to communication and management skills， improve later in life.

This reflects centuries of evolution. In almost any society，age is associated with wisdom， experience， and growing social status. Youth was the time for learning the fundamentals of the profession that one would practice throughout adulthood. Once in that job， a worker would refine their skills as they gained experience， but they would probably not have to learn new competencies from scratch.

Today's training programs are ineffective partly because they usually target fluid intellectual abilities. For companies， the conclusion seems to be that retraining a workforce is too challenging， so when new skills are needed， it is better to pursue alternatives like automation，offshoring， and crowdsourcing.

The assumption that workers， regardless of their age and educational background， will independently do what it takes to keep up with technological change is a fallacy that risks creating an army of unemployed. Such an approach can be expected only of the most highly educated and qualified workers – those whose jobs are usually not even at risk from automation.

This may change in the future， because younger generations are growing up with the expectation of lifelong learning. But， in the meantime， policymakers should take steps to mitigate the complicated mental processes at the root of many people's professional inertia.

As we develop robots with increasingly human-like capabilities， we should take a closer look at our own. Only by learning to overcome – or at least evade –our cognitive limitations can we have long and fruitful careers in the new global economy.

随着新技术不断颠覆各行各业并接管曾由人类完成的工作，全世界的劳动者都为自己的未来感到担忧。但真正令人类无法在劳动力市场展开有力竞争的并不是机器人本身，而是我们自己的头脑，其存在种种心理偏见和认知局限。

在当今变化迅速的劳动力市场，那些最受欢迎的职业在五年或十年前甚至还不存在，例如数据科学家、应用程序开发员或云计算专家。据估计，在今天上小学的孩子中，有65%的人最终将从事现在还不存在的工作。

要想在这样的劳动力市场上取得成功，劳动者需要成为敏捷的终身学习者，乐于不断地适应，并愿意跨行业流动。如果一个职业遭到淘汰——这种变化几乎一夜之间就可能发生，劳动者就需要能够敏捷地转入另一个职业。

终身学习当能提供在具有活力的新行业出现时抓住其中机遇所需的智识灵活性和职业适应力，以及应对衰退行业中所出现冲击的恢复力。从道理上讲，培训中心只需确定未来公司需要哪些能力，并相应地设计课程。

然而，在欧元区，2017年只有约10%的劳动力接受过某种正式或非正式培训，并且这一比例随年龄增长急剧下降。如果终身学习是在劳动力市场展开竞争的关键，那为什么人们如此不愿这样做呢？

人们认为，终身学习在时间、金钱和精力方面需要付出极高的成本，而回报非常不确定，尤其是在技术频频造成颠覆的情况下。劳动者在失去工作或面临职业生涯的十字路口时常常会感到沮丧和绝望，这可能令上述观点得到加强。

人类在较年轻的时候就会经历认知能力衰退，与工作记忆、抽象推理和新知识处理有关的流动智力在20岁左右开始衰退。中年以后，这些能力大幅下降，从而令新技能的获取变得越来越具挑战性。只有与沟通和管理技能有关的固定认知能力会随年龄增长得到提升。

这反映了千百年的进化。在几乎任何一个社会中，年龄都与智慧、经验和不断提高的社会地位相关联。从前，青年时期是一个人学习与他整个成年阶段将会从事的职业有关的基本知识的时候。一旦开始从事那项工作，劳动者就会随着经验的积累提高自身技能，但他们可能不必从零开始学习新的能力。

如今的培训项目之所以效果不佳，原因之一是它们通常针对的是流动智力。对企业来说，结论似乎是：重新培训员工太具挑战性，因此，在需要新技能的时候不如另寻他策，比如自动化、离岸和众包等。

无论年龄和教育背景如何劳动者都会自主拼尽全力跟上技术变革的臆断是一种谬论，可能会带来一支失业大军。只有受教育程度和素质最高的劳动者会这样做，而那些人的工作岗位通常不受自动化威胁。

这在未来可能会发生改变，因为较年轻的一代又一代人从小就抱有对终身学习的预期。但与此同时，决策者应该采取措施缓和造成许多人职业惰性的复杂心理过程。

在我们研发能力与人类越来越像的机器人的同时，我们应该对我们自身的能力进行更仔细的审视。只有学会克服或者至少是避开认知局限，我们才能在新的全球经济中拥有漫长且富有成效的职业生涯。（李莎译自世界报业辛迪加网站7月4日文章）