Summary of Key Points
In an interview, Andrew Ng dispelled the pessimistic notion that AI will lead to an employment apocalypse. His core argument is that AI will not directly replace entire jobs but will restructure the tasks within them. While AI can handle 30%-40% of the execution-based tasks, the remaining 60%-70% of tasks that require judgment, initiative, and cross-domain skills will become even more valuable. He also pointed out that the outdated education system results in skills that do not match the needs of the job market, and that AI can lead to a "cognitive offloading" (the loss of the ability to think independently) in the learning process. Additionally, AI will broaden the boundaries of jobs, making it possible to work across different fields. However, what is more crucial than knowing how to use AI is the ability to identify problems and take action to solve them. Ng emphasized that AGI (Artificial General Intelligence) is still a long way off, so there is no need to worry excessively.
Detailed Interpretation of Key Points
1. Don't worry about being replaced by AI; break down your tasks first
Ng advised against focusing on whether AI can replace your entire profession. Instead, break down your job into smaller tasks. For example, software engineers: while AI is capable of writing basic code, excellent engineers are actually busier as they delegate the repetitive and simple tasks to AI and focus on the more complex aspects such as understanding business requirements, communicating with the team, making critical decisions, and taking on project responsibilities. The programmers who still write code using traditional methods are the ones at risk.
Practical advice: Your job is not a single, monolithic task but rather a set of interconnecting tasks. AI will handle the simpler parts, leaving the more complex, thought-intensive, and responsibility-heavy tasks for you to handle.
2. Is it hard for young people to find jobs? Maybe the skills taught in schools are outdated?
The host mentioned that young people are more vulnerable to the impact of AI, but Ng believes the problem lies with the education system. University courses take too long to update. By the time new skills are taught, they may have become obsolete by the time students learn them. He suggested that students should not rely solely on school courses and should learn to work with AI, using it to handle its capabilities and focusing on tasks that it cannot handle themselves (such as understanding problems and being accountable for results). The interns in his team, including high school students, are highly efficient because they use AI as a tool while actively participating in the core aspects of their work.
3. AI can help you with tasks, but be careful not to rely too much on it
Ng noted that AI is not an ideal learning tool because it tends to complete tasks for you. For example, students may use AI to complete assignments, which may temporarily improve their grades, but they may not retain the knowledge because the process of thinking is omitted. He experienced this himself, forgetting how to solve problems after using AI to assist him.
Key reminder: Using AI to improve efficiency in work is fine, but don't let it provide you with the answers directly. What you need is to engage your brain in the difficult reasoning process, as this is what truly leads to learning.
4. AI is broadening job boundaries; you need to be able to work across fields
In the past, front-end and back-end development were separate fields, but AI has made it easier to switch between them. This trend will extend to other roles such as marketing and HR. For instance, marketers can now use AI to automate tasks throughout the marketing process, and recruiters can handle the entire recruitment process from finding candidates to following up on interviews.
Career development advice: The focus should not be on mastering existing skills to perfection but on acquiring skills that connect different areas of work. For example, marketers should learn to use AI to create small tools, and HR professionals should learn to automate processes. This does not mean changing careers; it means using AI to simplify business tasks.
5. Being proactive in identifying and solving problems is more important than just using AI
When hiring, Ng looks for candidates who can take the initiative to build things rather than just know how to use AI tools. For example, his marketers can write code and create automated tools, and his team includes "recruitment engineers" who use their engineering skills to optimize recruitment processes. He predicts that future roles will combine elements from different fields (e.g., "marketing engineers" and "HR engineers").
Core ability: AI reduces the cost of building solutions, making it possible to address smaller, previously unattainable tasks. Therefore, being proactive—identifying problems, creating prototypes with AI, and proving their value—is more important than just knowing how to use AI. In large companies, those who stick to their assigned tasks will become less valuable, while those who take the initiative will be more sought after.
Final Summary
AI is not here to take your jobs but to free you from repetitive tasks so you can focus on more valuable and human-intensive tasks such as judgment, communication, and problem-solving. Don't worry; start by trying to use AI to solve a small problem yourself.