### — Deep Reflections from a Family Educator and Parent
#### Starting with a Viral Article Recently, the topic "Liberal arts students are the biggest winners in the AI era" has been going viral. NVIDIA founder Jensen Huang's views have added fuel to the fire—he publicly stated that children in the future don't need to learn programming because human language is the programming language of the future; life sciences and humanities are the most difficult and most worthwhile fields to invest in. Qihoo 360's Zhou Hongyi also remarked that in the AI era, liberal arts students will be more sought after than science students.
These viewpoints are highly viral and have indeed touched a nerve for many parents. But as the President of the Shenzhen Family Education Association and a father of two, I want to say: What truly deserves our reflection is not whether to study "liberal arts or science," but a more fundamental question: What should we teach our children in the AI era?
#### Three Fundamental Changes Taking Place I am an educator, a parent, and someone who is personally practicing AI collaboration. Together with AI, I developed a WeChat mini-program called "Chuanjia" (Family Heritage)—a digital platform that helps families record collective memories, pass down family traditions and instructions, and co-build family funds. I don't know how to write code, but I am responsible for all product decisions, while AI handles all the coding. This project has made me deeply realize that AI is changing three things simultaneously:
First, the learning method has changed. Learning in the past was about "remembering answers." Teachers taught, students memorized, and tests examined. The underlying assumption was that the human brain is the best knowledge container. But now, anyone can obtain a PhD-level answer within three seconds. When "remembering what" is no longer scarce, "asking what" becomes the true differentiator. Jensen Huang is right; human language is indeed becoming the "programming language." The way you ask AI, the precision of your requirements, and the depth of your follow-ups directly determine the quality of the value AI returns to you. This is not a victory for liberal arts; it's a victory for those who can think and express themselves.
Second, the production method has changed. My experience with the "Chuanjia" project is the best illustration. Three years ago, it was a fantasy for someone without a technical background to build a functional digital product from scratch. But now, as AI lowers the threshold of "execution" to nearly zero, this is where the real value in Jensen Huang and Zhou Hongyi's views lies: people who were previously barred by technical thresholds—those good at insight into needs, understanding human nature, and grasping direction—have suddenly gained the opportunity to compete with technical talent. But note, this is not a "spring for liberal arts students," but a "spring for people with ideas." Liberal arts students without ideas will still be eliminated.
Third, the collaboration method has changed. Future work modes will no longer be just "people managing people," but "people managing AI" and "people collaborating with AI through AI." Everyone might simultaneously direct multiple AI assistants to complete different tasks. This means you need to be like a director—understanding the big picture while giving precise instructions to each "actor." What is this ability, ultimately? It is a competition of core underlying abilities.
#### Five Truly Valuable Core Abilities Based on my research in family education, combined with my personal experience raising children and building AI products, I believe these five abilities are the true "hard currency" of the AI era:
1. The Power of Questioning Einstein once said that if he had an hour to solve a problem, he would spend 55 minutes defining it. In the AI era, this sentence carries more weight than ever. Those who can ask good questions can derive ten times more value from AI than the average person. In the process of developing "Chuanjia," my biggest realization was that for the same functional requirement, different descriptions lead AI to provide vastly different solutions.
2. Aesthetic Ability AI can generate a thousand design schemes, write ten thousand poems, or compose a hundred stories in a minute. But the judgment of "which one is good?" still relies on humans. Aesthetics is not a luxury for the rich; it is the core filtering ability of the AI era. A person with aesthetic ability can pick the truly valuable content from the sea of AI output, while those without it will be overwhelmed by AI's mediocrity.
3. Empathy The smarter AI gets, the more it needs humans who truly understand "humanity" itself—understanding emotions, motivations, and the unquantifiable things behind the data. Face-to-face conversations, hugs, and reconciliation after an argument in the family—these are the true soil for empathy.
4. Integration Ability AI is good at handling local problems but not at global judgment. This cross-domain integrated judgment is an area where humans still lead by a wide margin. A child who understands both music and programming, or who reads history while following technology, will have far more competitiveness in the AI era than a "single-specialization" individual.
5. Expression Ability In the AI era, expression ability has at least two meanings: first, "making AI understand you"—the more precise and structured your description, the better AI's return; second, "making people willing to listen to you"—in a world where AI can help everyone write 80-score content, those who can produce 95-score expression will be exceptionally scarce.
#### Three Words from My Heart as a Father Having talked so much about "abilities," I want to say three words from the bottom of my heart as a father:
First: Don't chase the trends, chase the core. Specific professional knowledge might become obsolete in five years, but questioning, aesthetics, empathy, integration, and expression—the shelf life of these core abilities is a lifetime.
Second: Let children contact AI as early as possible, but don't let AI replace thinking. AI is the best learning partner but the worst stand-in. Let children use AI as a "super tool," not as an "all-purpose outsource."
Third: The more the technology races ahead, the more we must return to the essence of education. Technology changes, tools change, but human needs for love, understanding, belonging, and meaning have never changed. These things are not taught in school or given by AI; they are "soaked" in the family.
Recording growth, passing down family traditions. This is something AI can help with, but only family members can complete.
--- Author: SKYZOU President of Shenzhen Family Education Association Founder of "Chuanjia" Family Growth Community Platform