Partnering technology: As AI agents take on more routine tasks, there will be a corresponding need for humans to focus on planning, checking and directing.
As we welcome 2026, let’s take a moment to consider how far artificial intelligence (AI) has come. It was only three years ago that the introduction of large-scale generative AI took the world by storm. To some, it was magic; to others, a threat.
We were spending time chatting with AI chatbots, amazed by a software that could respond to our queries and even generate computer code. Today, AI, particularly generative AI, has become commonplace in our lives. Simple chatbots have evolved to autonomous agents and we have to view ourselves not only as users of AI but also as partners of technology in the new workforce.
Fear of replacement
Let’s talk about the elephant in the room. In the many presentations I have given on AI, the question I get asked most is: “Will jobs get lost because of AI?” This is a valid concern.
Recent research indicates that up to 57% of work hours in developed economies can be automated with current technology. This is an alarming prediction.
This, however, does not mean that 57% of jobs are being eliminated. It only implies that 57% of the tasks that occupy our work hours are being transferred to AI.
Repetitive tasks such as scheduling, data entry, basic research, and writing emails are being delegated to AI. However, a job is more than a list of tasks; it is a role requiring judgement, ethics and human connection.
Think of the kind of work that radiologists do. When we started using AI for image analysis, many predicted the slowing demand for radiologists. However, employment for radiologists increased. This is because AI handled image analysis, allowing doctors to focus on complex decision- making and patient care; the work became more valuable, not less.
Looking ahead, roles aren’t vanishing; they are being reimagined. Research shows that more than 70% of today’s skills can be applied in both automatable and non-automatable work.
The accountant who knows tax law is still vital, but in the age of AI, they spend less time on the spreadsheet and more time advising the client on strategy while an AI agent handles the calculation.
Our new ‘coworker’
To remain relevant, we need to understand the technology underpinning this paradigm shift. Until now, we have been accustomed to technology that follows a set of rules or answers a question. Such technology is passive in nature.
Today, agentic AI is emerging. Such systems comprise virtual team members who can strategise and complete complex, multi-step tasks.
A chatbot is one thing; an agent is another. An agent acts. Think of a prioritisation agent that prioritises sales leads based on data analysis or a scheduling agent that handles our calendar and sets up appointments for us.
In software development, we are currently experiencing a disruption in which we use coding agents that can develop an entire application, while software engineers are left to verify the result.
This means there has been a change in the skillset needed to succeed in the age of AI. There is a rapid increase in the need for AI fluency. This is the ability to use AI tools and orchestrate the flow of data and outcomes across multiple tools and agents. The demand for these skills is growing faster than the demand for any other skill.
What robots can’t do
The more advanced technology gets, the more our non-technical, uniquely human talents become valuable. Those involving helping, caring, or managing others, though, are among the least likely to change.
An AI system can uncover a sales trend, but not negotiate a fragile deal with an unhappy client. It cannot coach a younger colleague who has confidence issues.
Thus, it is important to double down on what makes you human. Resolving human conflict, creative design thinking, and leadership are domains where computers have yet to see the light of day. Our soft skills have become our “hard currency”.
‘Skill of the year’
To future-proof one’s career, focus on the orchestration paradigm. As AI agents take on more routine tasks, there will be a corresponding need for humans to focus on planning, checking and directing. This demands a different level of critical thinking.
One must be able to audit the work done by one’s AI agents. One must be able to identify hallucinations or errors and offer contextual knowledge that AI systems lack.
For leaders and managers, our role is evolving as well. We are no longer leading just people; we are leading a hybrid workforce comprising humans, agents and robots. Great managers of tomorrow will be those who can redesign end-to-end workflows and processes involving human and machine interactions.
A guide for 2026
As we enter this new year, I urge students, parents and professionals to embrace a spirit of experimentation:
> Don’t just read about AI agents. Work with them. Take advantage of the tools you have access to. Mastery comes from experience. An experimental culture will give us the quickest education.
> Develop relevant skills. Keep in mind that skills like critical thinking, problem-solving, communication, collaboration and adaptability are universal skills. They apply to any sector.
> Be the human in the loop. With every use of AI, always remember that we provide the oversight and governance. A machine can give us the results, but humans drive the outcome and impact.
The future workforce is a partnership of people, AI agents and robots working together. The economic benefit is staggering – trillions of dollars worldwide – but achieving it means learning to dance with our new partners. Happy New Year, and here’s to a smarter, more human 2026.
Asia Pacific University of Technology & Innovation (APU) vice-chancellor Prof Dr Ho Chin Kuan is a fellow at the Overseas Chinese Development Research Center of the Yangtze Delta Region Institute of Tsinghua University, China. Prof Ho, whose interests include data science, AI, machine learning and complex systems, works with leading educators to co-build the future of edtech. The views expressed here are the writer’s own.
