Getting Started in AI Without Knowing How to Code

🤖 AI tools accessible to everyone (zero code)

Good news: the best AI tools in the world are available directly in your browser. No installation, no terminal, no code.

💬 ChatGPT (OpenAI)

ChatGPT is probably the most well-known AI tool in the world. And for good reason: it is incredibly versatile.

What you can do with ChatGPT (free):
- Write emails, letters, resumes, cover letters
- Summarize long documents
- Brainstorm ideas
- Learn any subject (it explains like a patient teacher)
- Translate text with context
- Analyze images (GPT-4o)
- Create images (integrated DALL-E)

Free vs paid version:

Tip: The free version is more than enough to get started. Upgrade to Plus only if you use it daily.

🟣 Claude (Anthropic)

Claude is ChatGPT's direct competitor, developed by Anthropic. It excels in certain specific areas.

Why choose Claude:
- 🏆 Better at writing: more natural texts, less "robotic"
- 📄 Giant context window: can analyze 200+ page documents at once
- 🎯 More accurate: fewer "hallucinations" (made-up information)
- 💻 Artifacts: creates documents, code, and interactive visualizations

What Claude does better than ChatGPT:
- Analysis of long documents (reports, contracts, books)
- Creative and professional writing
- Explanation of complex concepts
- Following nuanced instructions

Free vs Pro version:

🔷 Google Gemini

Gemini is Google's AI, directly integrated into the Google ecosystem (Gmail, Docs, Search).

Gemini's strong points:
- 🌐 Google integration: works natively in Gmail, Docs, Sheets
- 🔍 Real-time web access: always up-to-date information
- 🖼️ Multimodal: understands text, images, video, audio
- 💰 Free: very generous base version

Ideal use cases:
- Searching for up-to-date information
- Summarizing emails in Gmail
- Writing assistance in Google Docs
- Analyzing Google Sheets spreadsheets

🔍 Perplexity

Perplexity is not a classic chatbot: it's an AI search engine. It answers your questions by citing its sources.

Why Perplexity is unique:
- ✅ Cited sources: every answer includes links to the sources
- ✅ Fresh information: real-time web access
- ✅ No hallucination: verifies its answers
- ✅ Clean interface: simpler than Google for complex questions

When to use Perplexity instead of ChatGPT:
- When you need verifiable facts with sources
- For research (articles, studies, statistics)
- When the information needs to be recent (news, prices, events)

📊 Comparison table of the 4 tools

Our recommendation for getting started: Start with ChatGPT (the most versatile) and Perplexity (for research). Try Claude when you need high-quality writing or to analyze long documents. To help you choose, you can check out our detailed comparison Claude 4 vs GPT-5 vs Gemini 3 : le comparatif honnête que personne ne fait.

🛠️ No-code platforms to go further

The tools above are perfect for one-off interactions. But if you want to automate tasks or create workflows, you need no-code platforms.

OpenClaw: your personal AI agent

OpenClaw is a unique solution: a personal AI agent that you control from Telegram. No need to code — you talk to it in natural language and it executes.

What OpenClaw can do for you (without code):
- 📱 Answer your questions via Telegram, 24/7
- 📝 Write content (articles, posts, emails)
- 🌐 Search the web and summarize the results
- 📊 Analyze files and data
- 🔄 Automate recurring tasks
- 🖼️ Analyze images you send it

How to get started with OpenClaw:

1. Ask someone technical to install it on a VPS (20 minutes)
2. Set it up with your Telegram account
3. Talk to it like a human assistant!

The interaction happens entirely in natural language: you send a URL to summarize, a request to write a LinkedIn post, or a text to translate, and the agent replies directly in your messaging app with the formatted result.

Unique advantage: Unlike ChatGPT or Claude (limited to a conversation), OpenClaw is an agent — it can execute actions, access files, browse the web, and chain complex tasks automatically. If this concept interests you, we have detailed how to create an AI agent that works 24/7 for €29/month.

⚡ Zapier AI: connect your apps

Zapier is the leading platform for connecting applications together. With its AI features, it's even more powerful.

Examples of Zapier + AI automations:
- New email received → AI summarizes → Slack notification
- New blog post → AI generates social posts → auto-publish
- Form filled out → AI analyzes → personalized email response
- Twitter mention → AI assesses sentiment → alert if negative

How it works:

1. Trigger: An event triggers the workflow (email received, form filled out...)
2. AI Action: Zapier sends the content to an LLM (built-in ChatGPT)
3. Final Action: The result is sent somewhere (email, Slack, Google Sheets...)

Price: Free up to 100 tasks/month. Paid plans starting at $19.99/month.

Limits: Zapier's built-in AI uses basic models. For quality, you need to connect your own API key.

Make (formerly Integromat): the visual alternative

Make is similar to Zapier but with a more powerful visual interface. You build your automations by connecting blocks like Legos.

Advantages of Make vs Zapier:
- More intuitive visual interface (drag & drop)
- More control over the data flow
- Better value for money (more generous free plan)
- Native AI modules (OpenAI, Claude)

Concrete example: no-code content pipeline with Make:

The process takes place in several visual steps. First, a Google Sheets spreadsheet serves as a source of topics. An AI module then generates the article from this data, followed by a second AI module that optimizes the SEO. In parallel, another module translates the content into English. Everything is saved in Notion, then sent as a draft to WordPress. All of this is built visually, without a single line of code.

📊 Comparison of no-code platforms

✋ 5 things to do starting today

Enough theory. Here are 5 concrete actions you can take now, each in less than 30 minutes.

1. Create your account on the 4 free tools (10 min)

Open these 4 links and create a free account on each one:

1. ChatGPT: chat.openai.com
2. Claude: claude.ai
3. Gemini: gemini.google.com
4. Perplexity: perplexity.ai

Why all 4? Each has its own strengths. By testing them all, you'll know which one to use for what. It's like having 4 specialized assistants instead of just one.

2. Ask the same question to all 4 tools (10 min)

Take a question you're genuinely interested in and ask it to all 4 tools. Compare the answers.

Examples of questions to test:
- "Explain [topic in your field] to me as if I were 10 years old"
- "What are the 2025 trends in [your industry]?"
- "Help me write an email for [specific situation]"

You will immediately see the differences in style, accuracy, and relevance. It's the best way to choose your favorite tool.

3. Have AI write a useful document for you (15 min)

Choose a document you actually need to write and ask AI to help you:

• An important email you've been putting off for days
• A meeting summary
• A LinkedIn post for your profile
• A job description
• A project plan

Effective prompt structure: Start by defining the AI's role (an expert in your field), then specify the type of document you want. Add the context of your situation, the desired tone (professional, casual, formal), the expected length, and list the specific points to include.

Key tip: The more context you provide, the better the result. Don't just say "write an email," say "write a professional email to my manager to ask for a raise, mentioning my 3 latest achievements: [X, Y, Z]".

4. Summarize a long document (10 min)

Find a long document you've never had the time to read:
- A PDF report
- A 5000+ word article
- A contract
- Class notes

Upload it into Claude (best for long documents) and ask for the following: the 5 key points with a two-sentence explanation for each, the important figures mentioned, the takeaways, and what might directly concern you based on your role.

Result: In 30 seconds, you'll have a structured summary of a document that would have taken you 1 hour to read.

5. Create your first "personalized assistant" (15 min)

On ChatGPT, create a custom GPT (or use the "Custom instructions" feature):

Recommended customization: Describe your professional profile (job, industry, technical level), specify that you communicate in French, and define your response preferences. For example, ask for clear and concrete answers, practical examples from your field, no technical jargon without explanation, and a suitable tone (professional, friendly, or direct).

With these instructions, every conversation will automatically be tailored to your profile. No need to repeat the context every time.

📈 Progressive learning path

Now that you have the basics, here is your roadmap for the coming weeks.

Week 1-2: The fundamentals

Objective: Use AI on a daily basis for simple tasks.

Skills acquired: Basic prompt, choosing the tool based on the need, daily time saved.

Week 3-4: The art of the prompt

Objective: Get 10x better results with structured prompts.

The 5 golden rules of prompting:

1. Role: "You are a digital marketing expert with 10 years of experience"
2. Context: "I am launching a B2B SaaS startup in fintech"
3. Task: "Write a content strategy for the next 3 months"
4. Format: "Present in a table format with columns: week, topic, channel, KPI"
5. Constraints: "Budget: €500/month. Team: 1 person. Target: CFOs"

Difference between a beginner and an advanced prompt:

A beginner prompt simply asks for LinkedIn post ideas. An advanced prompt first defines a specific role (LinkedIn expert with 50K+ followers), describes your profession and sector, details your target audience, and then asks for 5 post ideas for the week, specifying for each one: the hook (catchy first line), the structure (storytelling, list, or question), the call-to-action, the best time to post, and relevant hashtags (maximum 5). All of this in a professional but human tone, not corporate.

The difference in quality between these two approaches is huge. And it doesn't require any technical skills — just clarity in what you want.

Month 2: No-code automation

Objective: Automate your first recurring tasks.

This is the time to discover no-code platforms:

1. Week 5: Create your first Zapier or Make workflow

2. Ex: "When I receive an important email → AI summary → Slack notification"

3. Week 6: Explore an autonomous AI agent if you have access to a VPS

4. A complete AI agent controllable from Telegram

5. No code required, just natural language

6. Week 7-8: Combine tools

7. Ex: Perplexity for research → ChatGPT for writing → Zapier for publishing

To go further in this step, check out our guide on how to automate your business in 7 days with AI.

Month 3+: Specialization

Objective: Become an AI expert in your field.

Choose a specialization based on your profile:

If you fall into this last category, AI can even become a source of income. By the way, if you are considering creating a website to launch your business, we have detailed how much a website costs in 2026: the real price (not the one from agencies).

🧠 Understanding AI without jargon

To use AI effectively, a few concepts are useful to understand. Promise, no math.

What are LLMs?

LLM = Large Language Model.

Imagine a tool that has read the entire internet (books, articles, conversations, code...) and extracted language patterns from it. When you ask it a question, it doesn't "search" for the answer — it generates the most likely continuation of your text, based on everything it has learned.

This is why:
- It can write in any style
- It knows almost every subject
- It can sometimes "make up" information (hallucinations)
- It doesn't "know" what happened after its training date

Tokens and context

LLMs don't read words but tokens (pieces of words). One French word = about 1.5 tokens.

The context window is the amount of text the model can process at one time:
- GPT-4o: ~128,000 tokens (~90,000 words)
- Claude Sonnet: ~200,000 tokens (~140,000 words)
- Gemini: ~1,000,000 tokens (~700,000 words)!

In practice: Claude can analyze an entire book in a single conversation.

Free vs. paid models

For the majority of no-code uses, free versions are more than enough.

Hallucinations: the trap to know about

LLMs can make up information with absolute confidence. This is their biggest flaw.

How to protect yourself:
1. Verify important facts (especially numbers, dates, quotes)
2. Use Perplexity when accuracy matters (it cites its sources)
3. Ask for sources: "Cite your sources for this information"
4. Cross-check tools: if ChatGPT and Claude say the same thing, it's more reliable
5. Beware of ultra-precise details: "On March 23, 1987, Professor X published..." → verify!

🌟 Testimonials: they use AI without coding

To show you just how accessible it really is, here are some concrete use cases from non-technical people.

Marie, freelance communications consultant

"I use Claude to write the first drafts of my client articles. What used to take me 3 hours now takes 45 minutes. I spend the saved time fine-tuning the strategy."

Her tools: Claude (writing), Perplexity (research), Canva (design)

Thomas, SME owner

"My AI agent sends me a summary of my emails every morning on Telegram. It also prepares the replies for me. All I have to do is review and send."

His tools: AI agent via Telegram, ChatGPT for brainstorming

Sophie, law student

"I have Claude analyze my legal cases. It doesn't replace my professor, but it explains the concepts I don't understand so clearly that it has become my main revision tool."

Her tools: Claude (document analysis), Perplexity (case law research)

⚠️ Common mistakes (and how to avoid them)

❓ FAQ

Will AI replace my job?
No, but a person who uses AI might replace a person who doesn't. AI is an amplification tool, not a substitution tool.

What data can I share safely?
Avoid personal data (phone numbers, addresses), passwords, confidential financial information, and sensitive corporate secrets. The rest is generally risk-free.

Do I need to pay to use AI effectively?
No. The free versions of ChatGPT, Claude, Gemini and Perplexity easily cover the needs of daily use. Paid subscriptions are only justified for intensive use or advanced features.

How long does it take to become comfortable with AI?
By following the path described in this article, you will be autonomous in 2 to 3 weeks with 15 to 30 minutes of daily practice.

Does AI work in French?
Yes, the main models (GPT-4o, Claude, Gemini) master French very well, both in comprehension and in writing.

📌 The essentials

• 4 free tools cover 90% of needs: ChatGPT (versatile), Claude (writing and long documents), Gemini (Google ecosystem), Perplexity (search with sources).
• The key to a good result is the prompt: define a role, provide context, specify the format and constraints.
• No-code automation (Zapier, Make) allows you to move from occasional use to recurring workflows, without coding.
• Always verify important facts: hallucinations are the main pitfall of LLMs.
• Start today: 10 minutes is enough to create an account and ask your first question.

🚀 Conclusion

You now have everything you need to start your AI journey. Here is the natural progression:

1. Today: Create your accounts and test the 5 actions above
2. This week: Integrate AI into a daily task
3. This month: Master the art of prompting
4. Next month: Explore no-code automation
5. In 3 months: You will be the "AI expert" in your circle 😄

AI is not a passing fad. It is a tool as fundamental as the Internet was. And just like the Internet, those who adopt it early have a considerable advantage. The good news? It's not too late — and it has never been easier to get started. 🚀

🧰 Tools mentioned in this article

Quick recap of all the tools mentioned, with their main use case and pricing:

To get started right away, also check out our 10 projets IA à réaliser ce week-end — perfect for moving from theory to practice.
```