“How long does it take to learn AI?”
It’s one of the most common questions we get from working professionals. And it’s the one with the most unhelpful answers on the internet — usually a vague “it depends” followed by a course sales pitch.
So let’s give you the honest version, with real timelines, for 2026.
The short answer: two weeks to be useful, six months to be dangerous, twelve months to lead. The long answer depends entirely on what you mean by “learn AI” — because that phrase covers everything from “I want to chat with ChatGPT without feeling lost” to “I want to set AI strategy for my company.” Those are wildly different goals with wildly different timelines.
This post breaks down the realistic timelines for each, the three variables that change the answer, and how to figure out where you actually are on the path.
Five timelines, by goal
Most “how long to learn AI” content skips the goal question. That’s the mistake. Here’s what realistic timelines look like when you name what you’re actually trying to do.
1. “I want to use ChatGPT, Claude, or Gemini confidently at work” — 2 to 4 weeks
This is the foundation level. You’re not trying to build anything. You want to write a good prompt, know which model to use for which task, and stop feeling like everyone else gets it and you don’t.
What this looks like in practice:
- Week 1: Learn the vocabulary (tokens, context windows, hallucinations, temperature). Get a feel for how the three big models differ. Try the same task in all three.
- Week 2: Practice prompt patterns — role prompts, chain-of-thought, few-shot examples. Notice when the model is wrong and why.
- Week 3: Build a personal workflow. Pick one real task (writing, research, summarising) and do it with AI for two weeks. Compare to your old process.
- Week 4: Refine. Build a small library of prompts you reuse. Decide which model is your default.
After a month, you’ll be faster and more confident than 80% of the people using these tools at work. The bar is genuinely low right now — most professionals are still on Week 1.
If this is your starting point, AI Foundations — Level 1 is built for exactly this. Seven modules, no code, lifetime access.
2. “I want AI to actually change how I work, every day” — 1 to 3 months
This is where AI stops being a toy and becomes a real tool. You’re not just chatting — you’re building repeatable systems around the model. Meeting notes that write themselves. Research that takes 20 minutes instead of 2 hours. Drafts that come out 80% done.
What this looks like:
- Month 1: Build three repeatable workflows. Pick three tasks you do weekly and rebuild each one around AI. Document the prompts, the inputs, the outputs.
- Month 2: Add a second tool to each workflow. Most professionals stop at one model — the next jump is using AI for the parts it does well (first drafts, summaries, synthesis) and keeping humans for the parts it doesn’t (final judgment, client relationships, anything regulated).
- Month 3: Measure. How much time did you actually save? Which workflows stuck? Which did you abandon? The honest measurement is what separates “I use AI” from “AI is part of how I work.”
This is the level where AI starts paying back the time you invested in learning it. People who stop at level 1 don’t get this payback — they’re still doing one-off queries.
3. “I want to build AI tools, not just use them” — 3 to 6 months
This is where most people should stop. You don’t need to be an engineer. You do need to understand how AI systems work: what an API is, what retrieval-augmented generation is, what an agent is, what MCP does. You don’t write the code — but you can brief engineers, evaluate vendors, and understand the architecture.
What this looks like:
- Months 1–2: Learn the systems vocabulary. APIs, vector databases, embeddings, agents, RAG, MCP. You don’t need to build any of these — you need to understand what they do and when each one is the right answer.
- Months 3–4: Use no-code AI tools (Zapier AI, Make, n8n, Retool) to build something real. A customer support assistant. A research pipeline. An internal Q&A bot. Ship one.
- Months 5–6: Get comfortable with technical conversations. Read AI engineering blogs. Sit in on a technical meeting and follow 80% of it. You don’t need to code, but you need to not be lost.
This is the level where AI stops being a personal productivity tool and starts being a thing you can lead projects around. If you’re a senior IC, a team lead, or a founder, this is the level that pays the most.
4. “I want to lead AI strategy for my team or company” — 6 to 12 months
This is the executive level. You’re not building tools. You’re deciding which tools to build, which to buy, how to govern the risks, how to measure ROI, and how to get the organisation to actually use the thing you’ve rolled out.
What this looks like:
- Months 1–3: Get your personal AI fluency to level 3 above. You can’t lead a strategy you don’t understand personally.
- Months 4–6: Learn the business frameworks. Use case identification, build vs. buy analysis, governance models, ROI measurement, change management. Read what other companies have done and what blew up.
- Months 7–9: Pick a real business problem and ship a real AI solution. Not a pilot. A real one, with a real user base, measured against a real baseline.
- Months 10–12: Build the operating model. Who owns AI in your org? What’s the budget? What’s the risk appetite? How do new hires get up to speed?
This level is genuinely hard, and most people underestimate it. The technical part is the easy part — the organisational part is where most AI strategies die. If you’re at this level, Executive AI Strategy — Level 5 covers the six briefs you’d otherwise have to assemble yourself: competitive opportunity, risk and governance, build vs. buy, operating model, and leading change.
5. “I want to be a recognised AI expert” — 12+ months
This is the level where you write, speak, or research on AI as part of your professional identity. It’s a different game from the other four — it’s about public credibility, not internal competence.
It takes longer than people expect. Twelve months is the floor. Most recognised practitioners in the field have been at it for 3–5 years.
If that’s your goal, this post won’t get you there — but the other four levels are the prerequisite.
The three variables that change every timeline
Even with a clear goal, the timeline varies a lot by person. Three variables matter most:
1. Time per week. The biggest predictor. Two hours a week gets you to level 1 in about a month. Ten hours a week gets you there in a week. Be honest about how much time you’ll actually invest.
2. Background. If you’ve worked with any kind of data, software, or technical systems before, level 3 comes faster. If you’re coming from a non-technical role, level 1 is the right place to start, no matter your seniority.
3. The quality of your practice. Reading about AI doesn’t build skill. Doing real work with AI does. The people who learn fastest are the ones who pick a real task and rebuild it around AI in week one — not the ones who finish a course first.
How to start in the next 30 minutes
If you’ve read this far and you’re not sure where to begin, here’s the shortest path:
- Pick one task you do every week at work. Writing, research, summarising, planning, meeting prep — anything.
- Open ChatGPT, Claude, or Gemini. Use the one your company already pays for if there is one.
- Spend 20 minutes trying to do that task with AI. Not perfectly. Just try.
- Notice where it helped and where it didn’t. That gap is your learning agenda for the next month.
That’s the loop. Use AI → notice what worked → learn what you don’t know → repeat.
If you want a structured starting point, AI Foundations — Level 1 takes about a week to complete and gets you to level 1 above with a vocabulary, a workflow, and a set of reusable prompts.
If you’re already past level 1 and you’re thinking about strategy, governance, or leading AI at your company, Executive AI Strategy — Level 5 is the next jump.
Both are self-paced and lifetime-access. Pick the level that matches where you actually are today, not where you wish you were.
The fastest way to learn AI is to use it for real work, every week, for the next three months. Everything else is a slower path to the same place.