Short answer
Using AI does not automatically raise income. Harry Lui sets out a personal AI value path from experience, judgement and workflow to results and trust, for Hong Kong working professionals and SME specialists.
When many people start using AI, their first impression is that work has become faster. Reports, emails, presentations and data organisation can all be done in less time.
But a few months later they may notice a problem: their workload has grown, yet their income and life have not clearly changed.
The reason is that efficiency only means you can finish existing work faster; whether the market will pay more depends on whether you can solve more important problems, take on more complete outcomes and build capabilities others cannot easily replace.
If AI only lets us cram more tasks into the same working day, it has not really improved our lives — it has just accelerated the busyness we already had.
Start not with tools but with a result you can sell
“Knowing how to use AI” is not a clear enough market value. What customers or employers really buy may be more accurate sales analysis, faster quotes, more complete training content, clearer operating reports, or a workflow that reduces errors.
To increase income, the first step is not listing how many tools you know, but answering: which kind of work problem am I familiar with? Who would pay to have it solved? What result can be seen within a reasonable time?
AI’s role is to amplify, organise and productise the experience you already have — not to conjure up an expertise nobody needs.
Turn personal experience into a five-layer value system
Layer one: experience
People who have worked for years in logistics, retail, sales, administration or professional services usually build up a great deal of situational knowledge: which signals indicate risk, what customers really care about, which step most often goes wrong. This knowledge often exists only in their heads, and even they may not be able to explain it fully.
Layer two: judgement
Using AI to review real cases, you can organise experience into a judgement framework: what information to look at, which questions to ask first, how to compare options, and when to escalate. This step turns “I have done this for many years” into expertise others can understand.
Layer three: workflow
Combine the judgement framework with templates, checklists, data sources and review methods, and you have a stable workflow. You no longer start from zero each time but can complete work in a more consistent way.
Layer four: results
The workflow must connect to deliverables the market will buy, such as a diagnostic, an improvement plan, a working workflow, a corporate training session or a monthly analysis service. Results need a clear scope and acceptance criteria; you cannot sell only the idea that “AI is amazing”.
Layer five: trust
By consistently turning real problems, methods, limitations and results into content, the market gradually learns what you are good at solving. Trust does not come from posting a lot every day; it accumulates through clear positioning, reliable delivery and verifiable cases.
Three practical AI income paths
The first is increasing your value in your current job. Choose a repetitive but important departmental workflow and propose a measurable improvement, such as shorter processing time, fewer omissions or better report quality. The point is to let management see operating results, not to show off tool features.
The second is building a professional service. Organise years of experience into diagnostic, design and coaching services, serving one clear type of customer and one specific problem first. The first version does not need to do everything, but it must genuinely deliver one small result.
The third is productising the service. When the same problem keeps recurring, interviews, analysis, delivery and follow-up can become a standard process, with AI assistants, knowledge bases or automation added step by step. Only then does income have the chance to move from “selling your own time” to “selling a repeatable capability”.
Personal AI value also needs governance
When using AI to organise customer, company or colleague data, first confirm the data is appropriate to use and remove unnecessary sensitive content. You must also verify AI-generated analysis and recommendations yourself; professional responsibility cannot be handed to a tool.
Another risk is pursuing too many directions at once to make money quickly. Every new tool looks like a new opportunity, but building real income requires continually understanding the same kind of customer, accumulating cases and improving delivery.
Rather than changing topic every week, choose a problem you have experience in, that customers feel strongly about, and whose result can be verified; complete your first real transaction, then decide whether to expand.
The ultimate goal is more choice
The highest value of AI for an individual is not doing ever more, but gradually reducing low-value repetitive work and building knowledge, methods, work and customer trust you can take with you.
When you have repeatable workflows, demonstrable results and clear market positioning, you no longer depend on a single job or a single source of income. You can choose to take on higher-value work at your current company, or develop consulting, training, product or partnership income.
Speed is just one capability. Growth in income and choice is the real measure of whether personal AI transformation has succeeded.
Practical supplement
One-page personal value outcome card
Validate the need and deliver on a small scope before considering expansion. Working faster does not guarantee increased income.
Your entries remain only on this page and will not be saved after you refresh or leave; use Copy worksheet to keep the text.
Discussion question
If you chose only one, which verifiable outcome are you most confident you could complete using your existing experience?
Smark Global perspective
We do not just teach AI tools; we focus on turning personal experience into real work results, professional services and lasting capability.
Learn more: Smark Global AI Academy →Source: Original commentary by Harry Lui
Last updated: 2026-10-02（first published: 2026-10-02）