AI Era Ability Self-Assessment: What Level Are You At, and Where Do You Go Next? ​

We · October 2026

This is the sixth article in the "Core Abilities for the AI Era" series. The first five covered the amplifier effect, taste and discernment, problem definition ability, systems thinking, and AI literacy. This article consolidates everything into a self-assessment table and three upgrade paths, helping you定位 where you are and what to do next.

In the first five articles we covered a lot — AI is an amplifier, taste matters, you need to find the right problems, you need to build systems, you need to know AI's boundaries.

But after reading all that, you might be asking: it all sounds right, but where am I actually? Where do I start?

This article is here to answer that question.

We've organized the four core abilities discussed earlier into a self-assessment table. You can spend 5 minutes going through it to get a rough sense of your level. Then we've provided three upgrade paths — beginner, advanced, expert — each with specific practice lists and timelines.

No need to feel anxious, no need to bite off too much. Find your position, and just start from the next step.

Ability Self-Assessment Table: Four Dimensions × Four Levels ​

A quick note: this isn't a rigorous psychological assessment — it's just a quick定位 tool. For each dimension, pick the description that best fits you, no need to overthink it.

Dimension One: Taste and Discernment ​

Can you judge whether AI output is good quality, why it's good, and how to improve it?

Dimension Two: Problem Definition Ability ​

Can you turn vague needs into clear problems and find the real root cause?

Dimension Three: Systems Thinking ​

Can you combine multiple AI tools and steps into a complete workflow?

Dimension Four: AI Literacy ​

Do you know what AI can do, what it can't do, and how it should be used?

How to Use This Table ​

Step One: Score Yourself ​

Four dimensions, for each one pick your current level. Doesn't need to be precise, roughly right is fine.

For example:

• Taste: L2 (can tell good from bad, but can't explain why)
• Problem Definition: L2 (uses AI when there's a need, but it's scattered)
• Systems Thinking: L1 (still using tools individually, manually)
• AI Literacy: L2 (knows AI makes mistakes, checks important things)

Most people's distribution is roughly like this — some dimensions higher, some lower, totally normal.

Step Two: Find Your "Lowest Dimension" ​

Of the four dimensions, the lowest one is what you should work on first.

Why? Because AI is an amplifier — your overall effectiveness depends on the shortest plank.

For example, if your taste is great (L3) but your AI literacy is poor (L1) — what you produce might look great, but you might accidentally send sensitive data to an online AI service and end up with big problems.

Another example: if your systems thinking is strong (L3) but your problem definition is weak (L1) — you build a complex automated workflow, but it solves a problem that doesn't matter at all. The more efficient you are, the more you waste.

Fix the short plank first, then lengthen the long plank.

Step Three: Choose an Upgrade Path ​

Below we've provided three paths — beginner path, advanced path, expert path. Choose one based on your overall level and follow along.

Path One: Beginner Route (Overall L1-L2) ​

If most of your dimensions are at L1 or L2, this path is for you.

Goal: Upgrade from "can use AI" to "uses AI well."

Month 1: Build AI Literacy + Improve Judgment ​

This month's focus: build a clear understanding of AI, develop the habit of "not believing blindly, always verifying."

Specific practices:

1. Verification habit building — After each time you generate content with AI, spend 1 minute checking factual information (numbers, citations, examples). Make it a habit.
2. Understand where your data goes — Pick 1-2 AI services you use most, look at their privacy policies: will data be used for training, how long is it stored, can you delete it.
3. Taste comparison practice — Spend 10 minutes every day, find 5 good images and 5 mediocre ones, look at them side by side, try to say what's good about the good ones and what's bad about the bad ones.
4. Problem statement practice — Before using AI for anything, spend 2 minutes writing clearly: what problem am I solving, who is it for, what are the success criteria.

Expected outcome: Your understanding of AI will shift from "deifying" or "dismissing" to "clear-headed" — knowing what it can and can't do, and being able to more accurately judge output quality.

Month 2: Problem Definition + Basic Processes ​

This month's focus: upgrade from "scattered AI use" to "purposeful AI use."

Specific practices:

1. Five Whys exercise — Each week pick one problem you want to solve with AI, ask "why" five times in a row to find the root cause.
2. Subtraction practice — Every time you want to use AI, first ask yourself: if there were no AI, how would this problem be solved? Would the cost be high?
3. Draw flowcharts — Pick something you do regularly (like writing an article, making a video), draw out the complete process.
4. Find repetitive steps — In the flowchart you drew, find the 1-2 most repetitive, most time-consuming steps, and try automating them with AI.

Expected outcome: You'll start shifting from "I use whatever AI can do" to "I use AI for what I need." Usage efficiency noticeably improves.

Month 3: Your First Automated Workflow ​

This month's focus: build an automated workflow of your very own, hands-on.

Specific practices:

1. Pick a scenario — From your daily work or creation, pick a high-repetition, clear-steps scenario.
2. Build the workflow — Use ComfyUI (if doing image/video) or n8n (if doing business processes), chain together the steps of this scenario.
3. Iterate repeatedly — Run it several times, see where it gets stuck, where it's slow, where the效果 isn't good, keep optimizing.
4. Run locally once — If you haven't tried local deployment yet, install Ollama or ComfyUI, run a model locally, build that intuitive feel.

Expected outcome: You'll shift from a "using tools" mindset to a "building systems" mindset. And you'll discover — local deployment isn't that hard.

Path Two: Advanced Route (Overall L2-L3) ​

If most of your dimensions are at L2 or L3, this path is for you.

Goal: Upgrade from "uses AI well" to "uses AI systems to solve problems."

Month 1: Deepen Systems Thinking ​

This month's focus: upgrade from individual workflows to complete systems with multi-tool collaboration.

Specific practices:

1. Workflow assetization — Organize your commonly used workflows into templates, save them properly with documentation, instead of rebuilding every time.
2. Multi-model combination — Try using multiple models in one workflow — for example, using one model to generate first drafts, another for review and optimization, a third for style-specific output.
3. RAG入门 — If you haven't touched RAG yet, try building a simple knowledge base Q&A system with MaxKB or Dify. Doesn't need to be complex, just feel the power of "AI + private data."
4. Cost accounting — Calculate your current AI costs (time + money), then estimate what costs would be if you used local deployment. Do a comparison.

Month 2: Deepen Problem Definition ​

This month's focus: upgrade from "solving known problems" to "discovering the real problem."

Specific practices:

1. Redefinition exercise — Pick a problem you've been using AI to solve, force yourself to come up with at least three different ways to define it. Each definition corresponds to a completely different solution.
2. Root cause analysis — Pick a scenario where you feel "AI hasn't helped much," do a deep analysis: is AI not good enough, or is the problem definition wrong? Or is the data wrong?
3. Vertical domain deep dive — In the field you know best, deeply research what AI can do, what specialized models exist, what the best practices are.
4. Industry case studies — Find 3-5 successful AI cases in your industry, deeply research how they did it and why it worked.

Month 3: Local Deployment Deep Practice ​

This month's focus: upgrade from "trying local deployment" to "using local deployment to solve real problems."

Specific practices:

1. Pick a real scenario — Find a scenario you actually use daily, and do it with local deployment.
2. Model comparison — Try different models (different sizes, different architectures, different quantization levels), compare效果 and speed, find the best fit for your scenario.
3. Fine-tuning attempt — If you have the means, try fine-tuning a small model with LoRA. Don't worry about how good the效果 is, the point is understanding what fine-tuning is and what problems it solves.
4. Workflow migration — Migrate the workflows you do on online tools over to local. Compare the experience differences.

Expected outcome: You'll build a complete AI capability system — can judge problems, can design solutions, can execute and implement, can control risk. Both the quality and efficiency of your AI output will take a step up.

Path Three: Expert Route (Overall L3-L4) ​

If most of your dimensions are above L3, this path is for you.

Goal: Upgrade from "using AI systems" to "designing AI systems, creating new possibilities."

Months 1-2: Build Your AI System Stack ​

This stage's focus: build your own, fully locally controlled AI work environment.

Specific directions:

1. Complete local environment — From large models to image generation to voice to RAG, build a complete local AI environment. All data stays local.
2. Model library building — Build your own model library — which models are good for what scenarios, what are their characteristics, how to use them in combination.
3. Custom workflows — For your core business or creation scenarios, build deeply customized workflows. Not generic templates — tailor-made for you.
4. Automated operations — Automate every环节 of daily operations (content publishing, data monitoring, user feedback, quality detection) as much as possible.

Months 3-6: Create New Possibilities ​

This stage's focus: use AI to do things that simply couldn't be done before.

Not doing old things faster — doing things you never would have dared imagine before.

For example:

• One person building a complete product (from design to development to operations)
• One person running a media brand (content production + distribution + monetization)
• One person building a vertical AI service (industry-specific solutions)
• Using AI to explore entirely new creative forms or business models

There's no standard answer at this stage. Because people who reach L4 each go in unique directions — what problems you're most passionate about, which field you understand deepest, what you most want to create — that determines where you'll go.

But one thing is certain: when your AI ability is strong enough, the things you can do will far exceed what you previously imagined.

A Few Important Reminders ​

1. Don't Chase Perfection ​

Many people like to "wait until I'm ready before starting" — wait until I finish learning this, wait until I understand that, wait until I buy a better computer...

Don't. Find your current level, and just start from the next step. Every step forward brings its own rewards.

2. Short Plank First, But Don't Forget Your Long Plank ​

We said earlier to fix the short plank first — that's correct. But once the short plank is fixed to a certain degree, it's time to start lengthening the long plank.

What ultimately makes you stand out isn't your average level — it's how突出 your long plank is.

For example, if your taste is特别 good, keep going deep in that direction — in the AI era, the value of taste ability gets amplified. If you're especially good at systems thinking, go in the architect direction — people who can build systems are always scarce.

3. Compare With Yourself, Not With Others ​

This series mentions many impressive people — Pieter Levels making tens of millions a year alone, Midjourney doing $500M with 11 people... these are case studies, not goals.

Everyone's starting point is different, field is different, goals are different. Compare with yesterday's self — progress is enough.

AI is an amplifier — your starting point might not be high, but as long as you keep improving, AI will amplify your progress.

4. Doing It Yourself Is the Best Learning ​

Reading a hundred articles isn't as good as doing it once yourself.

Install a local model, build a workflow, do a small project — in the process of doing it hands-on, your understanding of AI will be completely different.

AI ability isn't learned — it's practiced.

Final Thought: Your Capability Boundary Is AI's Ceiling ​

Returning to the series' core proposition: AI is an amplifier.

AI can amplify your abilities tenfold, a hundredfold. But how well that amplification works depends on you.

• Poor taste → AI gives you a hundred images, and you can't pick the best one
• Poor problem definition → no matter how strong AI is, it can't help you solve the real problem
• Poor systems thinking → you only use tools scattered around, efficiency gains are limited
• Poor AI literacy → you might fall into pitfalls without even knowing how

But the flip side:

• Improved taste → AI can help you quickly produce high-quality work
• Accurate problem definition → AI can help you achieve order-of-magnitude efficiency gains
• Strong systems thinking → you can build systems that previously required a whole team
• Sufficient AI literacy → you can find the optimal balance between risk and reward

In the AI era, what's most valuable isn't how many AI tools you know how to use — it's your own foundational abilities.

Tools change, models iterate, technology progresses. But these foundational abilities — taste, judgment, problem definition, systems thinking, learning ability — they never go out of style.

Train these well, and no matter how AI changes, you'll stand firm.

Next article, also the final one in this series, we'll discuss the most core question: why is local deployment the litmus test for AI ability? Moving from "using AI" to "deploying AI" isn't just a technical upgrade — it's a leap in capability.

References:

• All citations and cases from the first five articles in this series, integrated into the ability framework of this article
• UNESCO AI Competency Frameworks — reference for AI literacy level classification
• "Extended Creativity: A Framework for Human-AI Creative Relations" (2025). arXiv:2506.10249. Support / Synergy / Symbiosis three modes (used for systems thinking level classification reference)
• McKinsey & BCG enterprise AI maturity models — industry reference for ability level classification