What Should Ordinary People Do in the Age of AI?
Amazon's profit surged 38% while laying off 30,000. Salesforce's AI handles 50% of customer interactions, eliminating 4,000 support jobs. This isn't a recession—companies simply don't need as many people anymore. What should we do in the next 5-10 years?
Amazon's 2025 profit hit $77.7 billion—up 38% year-over-year—while simultaneously laying off 30,000 people. Salesforce's CEO publicly stated that AI now handles roughly 50% of customer interactions, directly eliminating 4,000 support roles. IBM automated large portions of HR processes, cutting about 8,000 back-office positions.
These companies aren't facing an economic crisis. They simply don't need as many people anymore.
This post addresses one question: What should ordinary people do in the next 5-10 years?
How Is Value Created?
Before answering that, let's revisit a fundamental logic:
People have needs → Organizations produce → Output becomes products and services
When output value > input cost, profit emerges and social wealth accumulates.
As technology advances, production inputs keep falling while outputs keep rising. Many things that were "impossible" become possible.
Historically, New Tech Destroys Jobs—and Creates Them
Look at the Industrial Revolution: porters declined, farmers declined, armed escort services disappeared entirely. Yet total employment increased. New jobs fell into three categories:
Category 1: Cost reduction makes previously "uneconomical" demands viable
Before steamships, transoceanic shipping was too risky and slow to be economical. After steamships, costs plummeted, overseas trade exploded, and jobs like dock workers and international traders emerged.
Category 2: New production models require new management and support roles
Overseas trade created customs agencies, cargo insurance, and entirely new professions.
Category 3: Rising productivity unlocks higher-order human needs
Once survival is secured, people pursue meaning—therapists and counselors emerged because we could afford to care about mental health.
The core logic never changed: One party satisfies another party's need and receives compensation. AI lets us satisfy more needs, faster and better. Factory-specific management software, personalized education that was once an elite privilege—these previously uneconomical demands are becoming profitable. New opportunities follow.
But This AI Revolution Is Different
Historical optimism applies to society as a whole. For individuals, two critical questions demand honest answers:
Problem 1: Intensified Competition from Democratized Technology
Previously, having a skill or domain expertise was your long-term meal ticket. But AI dramatically lowers professional barriers—one person with AI can accomplish things they don't understand at all. Restaurant POS software? Soon anyone can build it, deeply customized.
Why should you do it better than others? Why should you earn that money? This is the first question you must answer.
Problem 2: Accelerating Pace of Technological Change
Major breakthroughs that once took a decade now happen in a year or months. How do you keep food on the table today while continuously learning enough to stay relevant and find bigger opportunities? This is the second question worth serious thought.
What Capabilities Do We Need?
To address these two problems, I believe three capabilities matter:
① Understanding people and their needs
When technology democratizes, similar products and services proliferate. Understanding users better and spotting needs earlier means capturing market rewards sooner.
② Building trust
AI can help you build products and serve users, but it can't build trust relationships for you. As competition intensifies, the answer to "why choose you?" increasingly isn't technology—it's relationships and trust.
③ Rapidly converting opportunities into action
Gaining information and mastering technology faster than others, then turning that lead into real delivery and income. The era of learning one skill and riding it to retirement may be over.
Common Questions
"Can I still be a programmer / novelist?"
Of course. But you'll compete with others, and ultimately those who use AI to write code or stories will win.
"What about delivery drivers or cleaners—AI can't touch those, right?"
These are realistic transitional options, but they shouldn't be the endpoint. Humanoid robots and low-altitude logistics may roll out slower than expected—these jobs could remain stable for 5-10 years. But on a 10-20 year horizon, nothing is certain. And as more people get squeezed out of other fields, these roles will flood with competitors, becoming brutal arenas of attrition. Treat them as transitions, not safe harbors.
So What Should You Actually Do?
I can't give you a "learn this skill and you're guaranteed safe" answer—technology moves too fast. A job I think is great today might be obsolete next week. We need to overcome our instinct for certainty and take initiative.
What I can offer are three directions:
① Observe and Think More
Stay current on technological developments and shifting landscapes. Keep asking yourself: "What's driving this? What might the implications be?" Find your own opportunities in the answers.
② Do More
Have an idea? Try it. With AI assistance, learning costs in any field have dropped dramatically, and launching online projects isn't expensive. Turn ideas you believe in into reality. Transform talent into monetizable systems. Build profitable brands. Create personal assets.
③ Speak More
Find places to express yourself. Train your ability to persuade. Let more people know you and trust you. The same product sold by different people yields vastly different results—that gap is real value.