One-sentence conclusion
If you want to improve yourself in your spare time, the most worthwhile priorities are “AI application capability + data analysis capability + automation capability.”
These capabilities are not only for programmers but also for roles in operations, marketing, sales, finance, administration, education, manufacturing, and consulting. Compared with simply learning a piece of software or chasing short-term hot side businesses, AI and data capabilities more easily create long-term compounding: they can improve work efficiency and expand career choices.
From the perspective of global skill trends, the World Economic Forum’s Future of Jobs Report 2025 lists AI and big data as one of the fastest-growing skill areas, with networks and cybersecurity and technological literacy also ranking high; at the same time, capabilities such as creative thinking, resilience, and lifelong learning are still rising. (World Economic Forum)
Why Is It More Advisable to Learn AI and Data Now?
In the past, the common answers to “improving yourself in spare time” were learning English, programming, writing, design, and personal finance. They are still valuable, but after 2026, what is more worth prioritizing is: whether you can apply AI, data, and automation tools to real work scenarios.
There are three main reasons.
1. Enterprises Are Widely Using AI, but People Who Can Truly Implement It Are Still Lacking
McKinsey’s 2025 global AI survey shows that 88% of respondents say their organizations already use AI regularly in at least one business function, but only about one-third of enterprises have begun to scale AI projects. (McKinsey & Company)
This means a real opportunity: enterprises need not only people who research large models, but also people who can apply AI to processes, products, content, data analysis, customer service, operations, and management.
2. Skills Update Faster, and It Is Increasingly Hard to Rely on a Single Experience for a Long Time
LinkedIn’s Work Change Report mentions that by 2030, the skills used in most jobs are expected to change by 70%, with AI being an important catalyst. (LinkedIn Newsroom)
Therefore, learning in spare time should not focus only on “mastering one more tool,” but on “improving the ability to adapt to new tools.”
3. AI and Data Are Not Single-Industry Skills but Cross-Industry Capabilities
Data from the U.S. Bureau of Labor Statistics shows that employment of data scientists is projected to grow 34% from 2024 to 2034, software development, quality assurance analysis, and testing roles are projected to grow 15% overall, and information security analysts are projected to grow 29%. Although these are U.S. data, they reflect long-term demand in digitalization, data, and automation directions. (Bureau of Labor Statistics)
In China, “AI + manufacturing” is also explicitly emphasized as an industry direction; related documents propose cultivating composite talent who understand both artificial intelligence and manufacturing applications, and improving AI literacy and skills across the workforce. (National Data Administration)
The 6 Most Worthwhile Capability Categories to Learn in Spare Time
1. AI Application Capability: Most Suitable for Ordinary People to Start With
Suitable for: All working professionals, especially non-technical roles, management roles, operations roles, content roles, sales roles, and administrative roles.
Core value: Improve efficiency, restructure workflows, and reduce repetitive work.
Learning content:
| Module | What to learn |
|---|---|
| AI tool use | Copywriting generation, information organization, meeting minutes, spreadsheet analysis, PPT outlines, customer service scripts, proposal generation |
| Prompt capability | Role setting, task decomposition, constraints, output format, iterative optimization |
| AI workflows | Transform “manual repeated operations” into “AI + tools + manual review” processes |
| AI risk awareness | Hallucinations, copyright, privacy, data security, result verification |
Why the prospects are good:
AI application capability is not “whether you can use a chatbot,” but whether you can apply AI to specific business. An operations person using AI for competitor analysis, a finance person using AI to assist with report interpretation, a salesperson using AI to generate customer profiles, and a teacher using AI to design course activities all belong to AI application capability.
For spare-time learning, it is recommended to start from “role scenarios” rather than learning deep learning algorithms from the start.
2. Data Analysis Capability: Needed by Almost All Roles
Suitable for: Operations, product, marketing, sales, finance, HR, management roles, and technical roles in transition.
Core value: Shift from “working by feeling” to “judging with data.”
Learning content:
| Level | Learning content |
|---|---|
| Introduction | Excel / spreadsheet modeling, data cleaning, pivot tables, common statistical metrics |
| Advanced | SQL, visualization dashboards, business metric systems, funnel analysis, retention analysis |
| Higher level | Python data analysis, machine learning fundamentals, predictive models, A/B testing |
Why the prospects are good:
Data analysis is not only needed by “data analysts.” Many roles in the future will require you to read data, interpret data, spot anomalies, and propose action recommendations. The U.S. Bureau of Labor Statistics’ occupational description of data scientists also emphasizes that their work includes collecting, categorizing, and analyzing data, building and testing models, and proposing business recommendations to stakeholders based on analysis. (Bureau of Labor Statistics)
3. Automation Capability: Free Up Time From Repetitive Work
Suitable for: Administration, operations, finance, customer service, sales support, project management, and individual entrepreneurs.
Core value: Raise the output ceiling and reduce mechanical labor.
Learning content:
| Direction | Example |
|---|---|
| Office automation | Automatically organize spreadsheets, generate daily and weekly reports, categorize emails, extract contract information |
| Process automation | Automatic notification after form submission, automatic distribution of customer leads, automatic task synchronization |
| AI automation | AI automatically summarizes materials, automatically generates drafts, automatically generates analysis reports |
| Light-code capability | Python basics, API basics, automation scripts, low-code tools |
Why the prospects are good:
The essence of automation capability is not “saving a few minutes,” but enabling one person to manage more processes, handle more information, and serve more business scenarios. For non-technical roles, automation is an important entry point for upgrading from “executor” to “process designer.”
4. Cybersecurity and Data Security Fundamentals: More Important as Digitalization Deepens
Suitable for: IT, operations and maintenance, development, managers, enterprise digitalization roles, and AI application roles.
Core value: Reduce data leakage, system risk, and compliance risk.
Learning content:
| Module | Learning content |
|---|---|
| Basic security | Password management, permission management, phishing identification, account security |
| Enterprise security | Data permissions, log auditing, interface security, backup and recovery |
| AI security | Sensitive data input risks, model output risks, prompt injection, content compliance |
| Career direction | Security analyst, security operations, data security, AI security governance |
Why the prospects are good:
Employment of information security analysts is projected to grow 29% from 2024 to 2034, and the U.S. Bureau of Labor Statistics also specifically notes that increased AI use and e-commerce growth will raise demand for security capabilities. (Bureau of Labor Statistics)
After AI tools are widely connected to enterprise processes, security awareness will change from an exclusive capability of IT departments into basic literacy for more roles.
5. Industry Knowledge + AI: Easiest to Create Differentiation
Suitable for: People who already have more than 2 years of industry experience.
Core value: Do not compete head-on with pure technical talent, but form a composite advantage of “understanding the business + understanding AI.”
Learning content:
| Original industry | Suggested AI capabilities to add |
|---|---|
| Education and training | AI lesson plans, learning analysis, intelligent grading, course design |
| Manufacturing | Visual inspection, predictive maintenance, industrial data analysis, intelligent production scheduling |
| Finance and accounting | Risk analysis, report automation, anomaly detection, compliance review |
| Marketing | User profiling, content generation, advertising analysis, private-domain automation |
| Human resources | Resume screening assistance, talent profiling, training content generation, performance data analysis |
| Healthcare | Medical text processing, health data analysis, compliant intelligent Q&A applications |
Why the prospects are good:
The State Council’s New Generation Artificial Intelligence Development Plan proposed relatively early to value composite talent development, including vertical composite talent who connect AI theory, methods, technology, products, and applications, and horizontal composite talent who master “AI +” in economics, society, management, standards, and law. (app.www.gov.cn)
In other words, what will be scarcer in the future may not be “people who only know AI,” but people who understand a certain industry and can apply AI in it.
6. Expression, Writing, and Structured Thinking: Even More Important in the AI Era
Suitable for: Everyone.
Core value: Improve communication, management, reporting, sales, and decision-making capabilities.
Learning content:
| Capability | Specific manifestation |
|---|---|
| Structured expression | Able to break complex problems into clear frameworks |
| Business writing | Able to write proposals, emails, reports, summaries, and SOPs |
| Questioning ability | Able to ask people questions and also ask AI questions |
| Judgment ability | Able to identify whether AI output is reliable |
| Review ability | Able to extract experience from results and continuously optimize |
AI can generate content, but it does not automatically guarantee that the content is correct, applicable, or compliant. Truly valuable people are those who can ask good questions, judge good answers, and organize good processes.
How Should Different Groups Choose?
| Your situation | Priority learning direction | What not to do at the start |
|---|---|---|
| Non-technical role, wants to improve efficiency | AI applications, automation, data analysis introduction | Diving directly into machine learning papers |
| Wants to switch to a digitalization role | SQL, Python, data analysis, AI application projects | Only learning theory without doing works |
| Already knows how to write code | LLM application development, RAG, Agent, AI engineering, security | Staying only at the level of calling tools |
| Works in operations/marketing/sales | AI content production, user analysis, automated growth, data dashboards | Only learning short-video editing techniques |
| Works in finance/administration/HR | Report automation, AI office work, process optimization, data governance | Only learning a single office software |
| Works in manufacturing/engineering | Industrial data, AI + manufacturing, visual inspection, predictive maintenance | Only learning general AI chat tools |
| Students or workplace newcomers | AI literacy, data analysis, basic programming, project works | Blindly signing up for too many courses |
90-Day Learning Path: An Executable Version for Spare Time
Days 1–30: Build AI and Data Foundations
The goal is not to “finish learning AI,” but to first be able to use it in daily work.
Suggested completion:
- Become familiar with the questioning, rewriting, summarization, analysis, and spreadsheet-assistance capabilities of common AI tools.
- Learn to use AI to generate outlines, reports, emails, and proposals, but manual proofreading is required.
- Review Excel / spreadsheet fundamentals, including functions, pivot tables, and data cleaning.
- Choose a real work scenario, such as “weekly sales data analysis” or “competitor material organization.”
Phased result:
Produce 1 AI-assisted work case, such as “Using AI to organize customer feedback and generate improvement suggestions.”
Days 31–60: Learn Data Analysis and Automation
At this stage, move from “being able to use AI” to “being able to design processes.”
Suggested completion:
- Learn SQL basics: querying, filtering, grouping, sorting, and aggregation.
- Learn Python basics: variables, lists, dictionaries, file processing, and simple data analysis.
- Learn to use AI for data interpretation, report generation, and anomaly detection.
- Design an automated process, such as “automatic organization of form data + AI-generated daily reports.”
Phased result:
Produce 1 small data dashboard or automation script.
Days 61–90: Do a Presentable Project
At this stage, form evidence rather than just learning records.
Suggested projects:
| Project direction | Example |
|---|---|
| AI office project | Automatically generate meeting minutes, task lists, and weekly reports |
| Data analysis project | Analyze data such as sales, users, inventory, courses, and advertising |
| AI customer service project | Build a Q&A assistant prototype based on FAQs |
| Content production project | Build an AI workflow for topic selection, writing, review, and publishing |
| Industry project | Create an AI application proposal based on your own industry |
Phased result:
Form a portfolio page or project description document that includes background, problem, method, tools, results, and review.
Do I Need a Certificate?
Certificates are not omnipotent, but they are valuable in three scenarios:
- You need a systematic learning path, and do not want to watch scattered tutorials.
- You need capability proof, for your resume, role transfer, job search, or internal promotion.
- You want to turn AI learning from an “interest” into “verifiable results”.
For example, CAIE (Saiyi) Registered Artificial Intelligence Engineer Certification provides foundation-level and expert-level certification for artificial intelligence engineering capability and suits people who want to systematically learn AI fundamentals, applications, and engineering practice. For non-CS learners, it is more advisable to start with the foundation-level capability framework and then accumulate works through real projects.
Please note: certification should serve capability building, not replace it. What is truly competitive is “certificate + projects + role scenario application.”
The Least Recommended Ways to Learn
1. Only Chasing Trends Without Looking at Your Own Role
Learning video editing today, livestreaming tomorrow, and prompts the day after easily ends up as “having touched everything but having nothing usable to solve problems.”
2. Starting With Overly Difficult Underlying Algorithms
If the goal is to improve workplace competitiveness in spare time, it is not advisable to start with deep learning mathematical derivations. A more realistic path is: first learn AI applications, then data analysis, then supplement programming and algorithms as needed.
3. Only Learning Tools Without Doing Projects
Tools change, but project capability does not easily become outdated. When learning AI, data, or automation, produce at least 2–3 demonstrable cases.
4. Treating AI as a Tool That Replaces Thinking
AI can improve efficiency, but it cannot take on judgment responsibility for you. Especially in scenarios such as finance, law, healthcare, education, and enterprise data, manual review and compliance awareness must be retained.
Most Recommended Learning Combinations
Ordinary Office Workers
AI applications + Excel/data analysis + automated processes + business writing
This combination is the most practical and can directly improve daily output.
People Who Want to Transition to Technical Directions
Python + SQL + data analysis + AI application development + project works
This combination suits transitioning to data analyst, AI application engineer, automation engineer, and junior development roles.
People With Existing Industry Experience
Industry knowledge + AI applications + data analysis + project management
This combination suits an “AI + industry” composite route, such as AI + education, AI + manufacturing, AI + finance, and AI + marketing.
People With Existing Technical Foundations
Large-model application development + RAG + Agent + MLOps + AI security
This combination suits development toward AI engineering, enterprise-level AI application development, and agent system design.
Conclusion
If you want to improve yourself in spare time, it is not advisable to ask only “which skill is hottest,” but to ask: which skill can continuously improve my work efficiency, career mobility, and future competitiveness?
Based on current trends, the priority can be ordered as follows:
- AI application capability: most suitable for ordinary people to start with and can quickly improve efficiency.
- Data analysis capability: helps you upgrade from an executor to a problem analyst.
- Automation capability: lets you amplify personal output with processes and tools.
- Cybersecurity and data security awareness: the deeper digitalization goes, the more important it becomes.
- Industry knowledge + AI: easiest to create personal differentiation.
- Expression, writing, and structured thinking: determine whether you can translate technical value into business value.
The truly promising learning direction is not learning a single tool in isolation, but forming a transferable capability combination: understanding the business, being able to use AI, being able to read data, being able to design processes, and being able to deliver results.
发表回复