Bored at work? Let AI do the heavy lifting

Ji Hang, 25, began working at a robotics company in Hangzhou city, Zhejiang province, a year ago. Instead of doing a stereotyped, repetitive, meticulous programming job, Ji has delegated much of the tedious coding work to artificial intelligence.

"The monotonous details of programming have been handed over to AI. It handles the syntax of various programming languages more proficiently than we do," said Ji, a graduate of an elite university in Zhejiang.

"We just need to provide the AI tool with a direction and teach it how to help us solve problems," he said, adding that he had witnessed significant advances in AI's programming capabilities, with today's version far better than what it was just six months ago.

Ji's experience reflects a broader trend across industries, where AI is taking over basic tasks.

As a result, entry-level job requirements have rapidly evolved. Newcomers are no longer just responsible for executing tasks. Instead, they are expected to undertake greater responsibilities, such as doing creative work and making judgments and decisions at an earlier stage of their careers, tasks which are considered more valuable, said industry experts.

The number of entry-level job positions has not decreased, they said, but the requirements are higher — AI literacy combined with creativity.

"The most valued skill in today's job market is the ability to integrate AI into real work processes to solve specific problems. In the future, the ability to effectively use AI as a work partner will significantly differentiate individuals in terms of work quality and efficiency," said Wang Xiangdao, founder and CEO of MoSeeker, a Shanghai-based enterprise providing digital recruiting solutions.

For Ji and his colleagues, it's not just the technical aspects of programming that are being rapidly overtaken by AI. Tasks like data annotation, which previously required dedicated personnel or even outsourcing, can now be handled automatically by large models. This approach is faster, highly accurate, and requires little proofreading.

"This shift has just occurred over the past year, and the change is quite evident," said Ji, who works at a company that develops autonomous exploration and mapping robots for places like underground mines that are hard for humans to reach.

Similarly, in the human resources industry, entry-level job tasks, such as searching for resumes online, will soon be completely replaced by AI. Leading companies now complete 80 percent of their initial recruitment processes through AI and automation, which offers significant advantages in terms of efficiency and fairness, Wang said.

Before, HR specialists had to search for talent online, review each resume, make comparative evaluations, arrange interviews, and follow up with candidates. However, AI is much faster, able to review tens of thousands of resumes received during a campus recruitment campaign and manage subsequent communications.

"Moreover, AI interviews are more impartial. Every human interviewer has personal preferences. They might give higher scores to candidates who are attractive or have backgrounds in big companies, but this doesn't necessarily reflect true ability," said Wang.

Over 2,000 libraries nationwide have deployed AI librarians, said Zhang Longqi, president of Chaoxing Group, a provider of digital and intelligent solutions for libraries, at a conference in Shanghai in May attended by hundreds of library managers from across the country.

"The accelerated implementation of intelligent technology and large model applications requires human librarians to collaborate with AI librarians to transform the vast resources of libraries into interactive, intelligent academic resources," he said.

"Human-machine collaboration should become the norm in library services for university faculty members and students."

A global report by the World Economic Forum on entry-level employees pointed out that most changes happening in the workplace today are due to task compression and role reshaping brought about by the advent of AI. This new technology is not simply taking away the first job, but redefining the content and value of entry-level positions, according to the report released in January.

Wang Wei, general manager of Talent Solutions Overseas Recruitment at LinkedIn China, said that AI is an inevitable trend in productivity upgrades, and young people have no choice but to adapt.

"Resisting or being annoyed by AI, essentially means falling behind the pace of the times and gradually being eliminated from the workplace," she said.

Balanced approach

On the Chinese mainland, 38 percent of enterprises are in the AI deployment phase, while another 38 percent are in the pilot phase, a report released in May by the global human resources service provider Aon Human Capital showed.

In addition, most Chinese enterprises said their employees were receiving AI-related training or skills enhancement, demonstrating that Chinese firms are more proactive than the global average in preparing talent for the future.

Ji said his employer values whether new recruits can work effectively with AI and use it in their work. The company hopes employees understand the core logic of using AI — to enhance efficiency and achieve a better work state — rather than relying on it completely without being aware of potential hidden risks.

Some HR experts said if newcomers are smart enough, they can use AI to make their work easier.

Ni Ke, an advertising professional in Shanghai, said tasks like summarizing the effectiveness of an ad campaign by watching hundreds of videos can be a challenge.

"However, AI can assist by analyzing the performance of each video ad, traffic conditions, and viewing duration. AI can even identify the top videos with the best performance and purchasing conversion effects, analyze what makes them successful and quantify these elements," said Ni, 34.

However, the same AI tools can yield vastly different outcomes depending on who uses them, experts said.

MoSeeker's Wang said the difference in applying AI tools correctly is the ability to instruct them how to deconstruct a problem.

"Knowing how to break down a complex problem into sub-problems, understanding the logical relationships between them, and the criteria for each sub-problem are crucial. AI can execute efficiently within a framework set by a user, but it cannot create the framework," Wang said.

More valuable roles

Jiang Jianghao, a new media video director, started her first full-time job this spring. With AI as a work assistant, newcomers in this field no longer start with information collection, but instead jump directly into the creative process, she said.

Behind each video project, extensive research is required, explained Jiang. Only after understanding all the content can ideas be extracted and creativity expressed.

AI has now replaced much of the painstaking work of information gathering. If a video is meant to promote a tourist destination, she can use AI to identify unique aspects and fun facts about the place. Jiang said she then needs to creatively integrate these elements with the desired content to showcase it in an engaging way.

"While AI can produce content, it cannot generate feelings. Truly touching works require human experiences, warmth, and judgment, which machines cannot imitate," said Jiang.

Many new employees shared this sentiment. With the arrival of AI, they are more often placed in roles where they must make judgments and decisions in complex real-world scenarios.

Ji said part of his work involves building a bridge between the "virtual and the real".

While large models can derive results through reasoning, whether these results can be implemented and meet expectations depends on humans using their practical experience and life knowledge, he said.

"In robotics, AI cannot make immersive judgments in real environments, and lacks lateral thinking and associative ability — unique human traits," said Ji.

Wang from LinkedIn said that the changes brought by AI have reshaped what constitutes valuable skills. This includes human judgment, making connections, and the ability to quickly adapt in complex environments, she said.

"For example, journalists today need to think about how to adapt to the AI era to excel in content creation," she said.

"Similarly, the legal profession is being reshaped by AI, as basic tasks like legal research and document drafting can now be handled by AI. Across various industries, the young talents moving fastest are those who quickly adapt to changes within their fields."

Human nature, go figure

Although AI is already beginning to replace some aspects of labor, many experts agree AI cannot replace the "soft skills" of humans, such as communication, collaboration, and face-to-face interaction.

Some consulting firms are now trying new methods to test whether candidates can do things that AI cannot. They include sensing subtle nuances during negotiations, and understanding another person's true concerns in ambiguous situations.

"While using AI to send emails will become increasingly common, complex and sensitive issues still require human-to-human communication," said MoSeeker's Wang.

LinkedIn data also showed 75 percent of companies agree that soft skills add the most value when AI handles mundane tasks.

Hybrid skills are becoming increasingly important as well, said Wang Wei of LinkedIn. A combination of technical fluency, human advantages, and adaptability will ultimately make job seekers more attractive to employers.

In May, the Shanghai Advanced Institute of Finance at Shanghai Jiao Tong University launched a part-time Master of Finance program. One of its aims is to cultivate hybrid talents with both AI technology understanding and financial business capabilities.

The institute's Ni Haiying explained that the program will be designed around real business scenarios in financial institutions to enhance students' problem-solving abilities regarding real financial issues.

"More than half of the program will be the practical part. We break down the required skill modules for real business problems, and integrate them into a solution-oriented curriculum," she said.

Ji and his colleagues spend part of their work time training large models. However, AI will by no means become equivalent to an individual person, he said. Instead, it will increasingly resemble a super-collective of all human knowledge.

"Even if one day robots can mimic everything humans see and touch, they cannot mimic human 'forgetting'. Human forgetting and thinking are random and uncertain, not based on fixed logic like equation derivation, whereas AI's decision-making relies on probability. This is the fundamental difference between machines and humans," Ji said.

zhouwenting@chinadaily.com.cn