AI Prompting Guide & Resource Hub
A beginner-friendly guide to AI prompts and chatbots, plus curated links to go deeper into ML.
Same AI model, wildly different results — the difference is almost always the prompt. This page covers what a prompt actually is, how to write better ones, which tools to try (and what they cost), and how "agentic" AI differs from a regular chatbot. If you're brand new to this, start here before diving into the deeper ML resources further down.
What Is a Prompt, and Why Does It Matter?
A prompt is just the instruction or question you give an AI model. But small changes in how you phrase that instruction can produce dramatically different answers — the model isn't reading your mind, it's responding to exactly what's on the page in front of it.
A vague prompt gets a vague, generic answer. A specific, well-structured prompt gets something you can actually use. That's the entire premise behind "prompt engineering" — it's less about tricks and more about communicating clearly.
Weak Prompt
"Write about home networking."
Too broad — the model has to guess your audience, length, and goal.
Stronger Prompt
"Write a 300-word beginner's guide to setting up a home VLAN, for someone who's only used a basic router before. Use plain language, avoid jargon, and end with one practical next step."
Specific audience, length, tone, and outcome — much less guesswork.
Prompting Principles That Actually Work
Be Specific
State the audience, tone, length, and goal. "Explain Docker" gets a generic wall of text; "Explain Docker to a home server beginner in 3 short paragraphs" gets something usable.
Give Context
Tell the model what you already know, what you've tried, or what constraints you're working under. The more relevant background it has, the less it has to guess.
Ask for Step-by-Step Reasoning
For anything involving logic, math, or troubleshooting, ask the model to work through it step by step. This tends to catch errors that a rushed, single-shot answer would miss.
Show an Example
If you want a specific format or style, show one example of it. Models are very good at matching a pattern once you've demonstrated it.
Specify the Format
Want a table, a bulleted list, a specific word count, or a certain structure? Say so directly instead of hoping the model guesses correctly.
Iterate, Don't Restart
If the first answer isn't quite right, refine it in the same conversation ("make it shorter," "less formal," "add an example") instead of starting over from scratch.
Popular AI Chat Tools
Where to actually try this stuff, and roughly what it costs. Pricing changes often, so treat these as a starting point and check each provider's site for current plans.
ChatGPT
Free + PaidOpenAI's assistant. Solid free tier; paid plans start around $20/mo, with higher tiers for heavier use.
Claude
Free + PaidAnthropic's assistant, this site's writing partner. Free tier available; Pro starts around $20/mo.
Gemini
Free + PaidGoogle's assistant, integrated with Workspace apps. Free tier available; paid plans start under $20/mo.
Perplexity
Free + PaidBuilt around search with cited sources — good for research. Free tier available; Pro around $20/mo.
Microsoft Copilot
Free + PaidBuilt into Windows and Microsoft 365. Free standalone version; deeper Word/Excel/Outlook integration requires a Microsoft 365 subscription.
DeepSeek
FreeFree web chat with a strong reputation for coding and reasoning tasks; usage-based API pricing for developers.
Beyond Chat: What Is "Agentic" AI?
A regular chatbot answers one message at a time — you ask, it responds, you read the result. An agent is different: you give it a goal, and it plans and carries out multiple steps on its own — searching the web, running code, editing files, or using other tools — checking its own progress along the way instead of waiting for you to prompt every single step.
Examples of agentic tools include coding assistants that can read a codebase, make changes across multiple files, and run tests on their own; browser agents that can navigate websites and fill out forms; and research agents that break a broad question into smaller searches and compile the findings into a report.
The prompting principles above still apply — being specific about the goal, constraints, and what "done" looks like matters even more once you're handing off multiple steps instead of just one.
Going Further: The ML Resource Hub
Curated links for anyone wanting to go beyond prompting and into building or studying AI/ML itself. Click a category to expand it.
Courses & Tutorials
Prompting & Generative AI
Classic ML & Deep Learning
Frameworks & Libraries
Core ML/DL
Agent Frameworks & Protocols
Vector Databases & RAG
Communities & Blogs
Frequently Asked Questions
Quick answers to common questions about AI prompts, chatbots, and agents.
What is prompt engineering?
Prompt engineering is the practice of writing clear, specific instructions to get better results from an AI model. It's less about tricks and more about communication: stating the audience, format, and goal explicitly instead of leaving the model to guess.
What makes a good AI prompt?
A good prompt is specific about the audience, tone, length, and desired outcome, gives relevant background the model wouldn't otherwise have, and asks for step-by-step reasoning on anything involving logic or troubleshooting. Showing one example of the format you want is often more effective than describing it in words.
What's the difference between a chatbot and an AI agent?
A chatbot responds to one message at a time and waits for the next prompt. An agent is given a goal and plans and carries out multiple steps on its own, such as searching the web, running code, or editing files, checking its own progress along the way instead of waiting for a prompt at every step.
Are AI chatbots like ChatGPT and Claude free to use?
Most major AI assistants, including ChatGPT, Claude, Gemini, and Perplexity, offer a usable free tier alongside paid plans that add higher usage limits and access to more capable models. Pricing changes fairly often, so it's worth checking each provider's current plans directly.
Do I need to know how to code to write effective AI prompts?
No. Prompt engineering for everyday use is a plain-language skill, not a programming one. Coding becomes relevant only once you move into building tools that call AI models programmatically through an API, which is a separate skill from writing effective prompts in a chat interface.
What is RAG (Retrieval-Augmented Generation)?
RAG is a technique where an AI model looks up relevant information from an external source, like a document database or vector database, before generating its answer. It's how AI tools stay grounded in specific, current, or private information instead of relying only on what they learned during training.
What is the Model Context Protocol (MCP)?
MCP is an open standard that lets AI models connect to external tools and data sources, like file systems, databases, or apps, in a consistent way. It's become one of the more significant developments in how AI agents actually take action rather than just generating text.