Where Do You Stand in the AI World? A 14-Tier Capability Ranking by Industry Impact

I built my own ranking of AI understanding and mastery: fourteen tiers from T0 to T13, ordered by a single axis, impact on the AI industry as a whole. The higher the tier, the broader the reach; the lower the tier, the more the impact narrows to one's own work. This table has no academic backing. It is purely my own observation, offered as a reference for locating yourself.

How I Developed This Framework

I started with only three tiers. The more I thought about it, the more granular it became, and it eventually expanded to fourteen.

The most critical decision was choosing the ordering axis. I tried ranking by technical depth, and I tried ranking by sensitivity to AI. In the end I settled on impact on the AI industry as a whole. The reason is simple: this axis has only one criterion. Whoever has a broader sphere of influence goes higher, with no ambiguity of the kind you get when someone ranks high on technical skill but low on commercial reach.

Technical depth and industry impact are broadly correlated. Building a foundation model from scratch is the hardest; distillation is next; fine-tuning follows; building applications on top of existing models is the lightest lift. But the two are not identical. That is why a team like DeepSeek, which distills an existing large model into a deployable smaller one, can rank above institutions that can produce a large model but cannot commercialize it, when the axis is impact rather than raw technical achievement.

Today's T0 May Not Always Be T0

I placed AI-native entrepreneurship and AI application tiers relatively low, and I want to be honest about why.

At this stage, everyone is still on the path of "innovating with AI." No decisive, clearly successful case has yet emerged. So right now, the people with the greatest influence on the industry as a whole are still the ones who built the foundation models. That is why T0 belongs to them today.

My own read is that as individual industries and domains progressively deepen their understanding of large models, the people who actually use AI to rewrite the world will climb tier by tier. The influence of model builders may still expand in absolute terms, but those who reshape the world with AI will expand even faster and gradually move above them. When that happens, the occupant of T0 may well be someone different.

The analogy that comes to mind is the railway. When railways first developed, the first wave of winners were the companies that manufactured rails and laid tracks. But once the full rail network was in place and trains started running and driving economic activity, the people who actually made the most money were those who understood how to build businesses around the flow of passengers. AI is still in the track-laying phase. The model builders are the first-wave winners. Once the infrastructure is complete, those who take these capabilities and reshape industries and create businesses will be the next, larger wave of winners. The people who profit most may well be those who know how to use models, not only those who build them.

The Full Ranking (T0 to T13)

• T0Model Creators (Commercially Viable): Can build a foundation model and bring it to market. Examples: Sam Altman at OpenAI, Dario Amodei at Anthropic.
• T1Large Model Distillers: May not have the resources to train the most capable foundation model from scratch, but can distill an existing large model into a smaller, controllable, deployable version. Done well (as DeepSeek has shown), the industry impact is substantial.
• T2Model Creators (Not Commercially Viable): Technically capable of building a large model, or a smaller-parameter model, but lack the ability or resources to commercialize it. Mostly institutions rather than individuals.

• T3Engineers at Commercially Viable Model Companies (Public Sharers): Work on models inside a commercially successful company, and also run a public channel or share on GitHub, making first-hand observations and learnings publicly available in real time. Their influence comes from this dissemination.
• T4AI Research Divisions at Large Corporations: Internal AI research units at major enterprises, focused primarily on published research or internal applications.
• T5University AI Research Labs: Academic labs engaged with the AI frontier. Deep understanding, but constrained by resources and scale.

• T6AI Application Framework Designers: Design the foundational tools others use to build applications: agent orchestration frameworks, RAG system architectures (the creators of things like LangChain and LlamaIndex). They do not train models themselves, but they define how the world builds with models.
• T7AI-Native Entrepreneurs (with Fine-Tuning Capability): Identify new value and business models in the AI era and build companies around them, with the added ability to fine-tune models.
• T8AI-Native Entrepreneurs (without Fine-Tuning Capability): Also entrepreneurial in the AI era, able to connect and apply others' models using RAG, agent frameworks, and similar off-the-shelf tools, but without fine-tuning their own.

• T9Enterprise AI Workflow Creators: Redesign organizational workflows from scratch in response to AI capabilities.
• T10Enterprise AI Workflow Adopters: Integrate AI into existing corporate processes to optimize them.
• T11Personal AI Workflow Creators: Design new personal workflows built around AI capabilities.
• T12Personal AI Workflow Adopters: Plug AI into their existing personal workflows to make them more efficient.
• T13Personal AI Users: Remain at the level of chatting, looking things up, and trying new features, without integrating AI into any workflow.

How to Use This Table

Here is how I use it myself:

1. Find roughly which tier you are at.
2. Look at what separates you from the tier above. That gap is your next direction.
3. There is no need to rush upward. Knowing your direction is enough.

Where I Place Myself

I position myself at T8, AI-Native Entrepreneur (without Fine-Tuning Capability). I can connect and apply others' models using RAG and agent frameworks, identify new value and business models, and build them into a business. But I do not fine-tune or train models myself.

I know there are many tiers above me. This table is also a reminder to myself: know where you stand, and know which direction to take next.

A Final Note

This is my own framework, not a proposed standard. The AI landscape moves fast. The ranking I have today may need revision in a few months. Take it as a reference point, use it to find roughly where you are, and let it prompt you to think about your next step. That is all it needs to do.