'A degree can get you in the door, but learning ability lets you stay and keep growing.'
Recently, in an interview, Fei-Fei Li publicly stated that at World Labs, when hiring software engineers, academic credentials matter far less than before — the company values learning outcomes, willingness to embrace AI tools, and the ability to grow fast, and will absolutely not hire practitioners who refuse AI collaboration tools.
In the AI age, what exactly does Fei-Fei Li mean by “learning ability”? How should we view the relationship between AI and our own growth?
01 The Core Competitiveness of the AI Age: Learning Ability Grounded in First Principles
Fei-Fei Li has never denied the value of degrees; what she opposes is the rigid mindset of “treating a degree as the finish line.” A degree is a “static achievement” that proves you have stage-by-stage learning ability and a basic cognitive framework (e.g., logical thinking, fundamental subject knowledge), and is the workplace's “entry ticket.” Without this foundation, even if you want to embrace AI collaboration tools, it is hard to grasp their underlying logic, let alone “use the tools to boost efficiency.”
Learning ability, by contrast, is a “dynamic capability” and the core engine for breaking through the degree ceiling. As knowledge depreciates faster than ever, the stock knowledge a degree represents quickly becomes obsolete; only learning ability lets us keep replenishing incremental knowledge, transform quickly amid job changes, and retain the initiative in using tools. The difference: a degree gets you in the door, but learning ability lets you stay and keep growing.
In the AI age, knowledge devalues fast and AI seems to know everything — so how should we keep learning? During his July 2025 visit to China, Jensen Huang offered this advice:
The core competitiveness of the AI age is learning ability driven by “first-principles thinking + critical thinking + value boundaries.” In the AI age, learning ability rests on first-principles thinking at its base, which demands we return to the essence of things and reason from there, not bound by fragmented information or AI's surface-level output; its core is critical thinking, the ability to maintain independent judgment on authoritative views and tool-generated content, distinguishing truth from falsehood and the boundaries of applicability; and the value boundary is to “uphold the true, the good, and the beautiful of the human world,” never using AI to fabricate false information or rejecting the use of AI for harm, always holding to the principle of “humans leading tools.” So in daily work, how should we collaborate with AI to improve our own learning ability?
02 Using AI to Upgrade Yourself: Don't Let Your Tool Make You Stupid
This learning ability is not a gift but a skill that can be deliberately practiced; it demands we proactively embrace new things, proactively break down problems, and proactively review and improve — far harder than passively receiving knowledge. Learning ability is neither simple knowledge accumulation nor the passive skill of using AI; it is a meta-ability that uses AI as an extension of thought to see through to essence, keep independent judgment, and hold the value baseline. In a Bloomberg interview this year, Fei-Fei Li offered this advice:
The First Principle of Applying AI: A Thinking Partner, Not a Super Search Engine
Most people use AI as a super search engine — for efficient retrieval of information and quick output of results — and over time this makes them neglect thinking and grow increasingly stupid. The first principle of AI is to be an extension of and resonance with human thought, a true “thinking partner”: it does not skip the steps of thinking for us, but helps us break down analytical dimensions, fill cognitive blind spots, and verify logical loopholes; it does not hand us the final conclusion but opens an entry point into thinking, with the human ultimately remaining in charge of decisions.
In the traditional era, “knowledge reserve set the ceiling,” a degree equaled a capability endorsement, and a single skill could sustain a long career; but in the AI age the logic of competition has been completely rewritten, shifting from “stock-knowledge contest” to the dual competition of “learning speed + AI collaboration.” AI can efficiently integrate and call on vast stock knowledge, so purely “knowledge-moving” or “mechanical execution” work is already in grave danger; what is truly scarce is the ability to “collaborate with AI to create incremental value” — e.g., using AI for material organizing while you do logical restructuring and innovative output, or using AI for scenario simulation while you make decisions and control risk.
03 Individuals Responding to Change: Rebuild Your Capability with AI
Fei-Fei Li stated bluntly that “AI replacing some jobs is an inevitable process” — it is the law of technological development. From the steam engine to electricity, from the personal computer to autonomous driving, every technological iteration brings job pains but ultimately gives rise to new employment landscapes. Debating merely “whether jobs increase or decrease” misses the point; the core of this transformation is “capability reconstruction,” not “unemployment panic.”
1. At the Individual Level: The “4+1” Practice Method for Cultivating Learning Ability
Build a first-principles thinking framework: focus on the core essence of your industry; when a problem arises, first reason back to essentials rather than be bound by appearances or AI's surface-level output;
Practice AI's first principle: always treat AI as a thinking partner, raising core doubts rather than demanding answers; use AI to break down dimensions and fill blind spots, then use critical thinking to verify logic and refine conclusions — e.g., after using AI to sort out industry trends, independently distill an action plan suited to yourself;
Build a “learning – collaboration – review” closed loop: set annual deep-learning goals (e.g., master a cross-domain skill, thoroughly learn a core tool), and construct a personal knowledge system; proactively use AI tools to amplify thinking efficiency (e.g., AI for material organizing, scenario simulation), and review collaboration results weekly to refine your method;
Strengthen continuous-iteration ability: follow AI's iterative developments, proactively learn new functions and scenarios, and try combining AI tools with your own work — e.g., a programmer uses AI to generate basic code then optimizes and adapts it; an office worker uses AI to draft a report framework then adds in-depth analysis;
Hold the value boundary: remember “don't let your tool make you stupid”; refuse to rely on AI for core thinking tasks (e.g., critical work analysis, professional content creation); never use AI to fabricate false information; and always stay clear-headed that “humans lead the tools.”
2. At the Organizational Level: Transformation Moves That Fit Learning Ability + AI Collaboration
Enterprises: break the “degree-above-all” hiring logic and build an evaluation system of “learning ability + AI collaboration logic + values” (e.g., assessing problem-solving approach and AI-tool ethics in interviews); provide AI-tool training and cross-department learning opportunities, guiding employees to correctly understand AI's first principle and avoid misusing the tools;
Education institutions: shift from “knowledge imparting” to “learning ability + AI collaboration + values training,” strengthen first-principles and critical thinking, and offer a course on “applying AI's first principle”; through case teaching, let students master the core ability to “think in concert with AI,” making clear the principle that “tools are an extension of thought, not a replacement.”
Technological change is irreversible and job replacement inevitable, but the real threat is not AI but “the self that stopped growing” and “the inertia of misusing tools.” The era waits for no one, yet it treats kindly those willing to grow proactively and use tools correctly. Once we truly master the learning ability to “think in concert with AI,” we gain the confidence to face all uncertainty: AI can replace the calling-up of stock knowledge and mechanically repetitive labor, but it cannot replace essence-thinking driven by first-principles thinking, cannot replace the independent judgment upheld by critical thinking, and still less can it replace the upright innovation of holding to values. Only when we master the learning ability to grow together with AI can we, in this era of accelerating knowledge depreciation and intensifying job change, shift from “passive adaptation” to “active leadership.”
