ChatGPT & Generative AI - The Complete Guide

Explore how chatbots like ChatGPT and Google Gemini work, how AI models are trained, and how inputs generate answers, including tool-using models, while understanding their capabilities and limitations.

Learn how ChatGPT functions as an application shell, validating input, managing visible and invisible chat history and system prompts, and passing context to the model via apps and APIs.

Explore how large language models, with billions of parameters in transformer-based neural networks, are trained on internet text to generate meaningful output.

Explain how text becomes tokens and token IDs for training, then maps to vector embeddings that encode token relationships. Show how the model predicts the token using probabilities and sampling.

Learn how foundation models are pre-trained to predict tokens and are fine-tuned with human conversations and synthetic data to become helpful assistant models.

Explore how reasoning models reveal a thought process before giving final answers. See how chain-of-thought prompting and training tune token prediction for better math and code execution.

Explore how context windows govern model predictions, handling long chat histories and inserted PDFs, and how modern models with larger windows enable summarization tasks that cheaper models struggle with.

Explore why these models hallucinate, how guardrails reduce it, and why outputs are only likely, not guaranteed to be correct; always double-check results in important tasks.

Explore how knowledge cutoff in model training limits recent information and can cause hallucinations, and compare fixes: cutoff-aware fine-tuning and giving the model web access to search for up-to-date data.

learn how large language models access tools through an application shell, where hidden prompts trigger tools like web search or code execution, guided by the system prompt.

Explore how large language models are created and how token prediction powers ChatGPT, recognizing hallucinations and knowledge cutoff, and using the application shell for prompt engineering and AI integration.

Explore core and advanced ChatGPT features, from prompt engineering and agents to AI models and the model picker. Learn to manage prompts, conversations, chat histories, and built-in tools with control.

Compare free and paid ChatGPT experiences, exploring how login status, model capability, and feature access differ, and learn to navigate the chat interface and start a new chat.

Navigate the ChatGPT interface, from the chat input box to file uploads and voice mode. Understand how model picker, temporary chat, and memory affect prompts and conversations.

Understand the model picker and versioning, from fast general models to reasoning models, and learn when to choose 4o, 4o mini, or O3 for complex prompts.

Discover how to craft concise, complete prompts for ChatGPT, use article text and delimiters, assign roles, and manage the context window with chat history.

Learn to edit prompts and fork chat histories to create multiple response branches, then regenerate messages with different models and web search to refine ChatGPT outputs.

Explore dictation and voice modes to interact with ChatGPT by speech, comparing transcription-based input with AI-generated voice replies for a natural, conversational experience.

Learn how to enable and use ChatGPT’s web search tool to fetch up-to-date information beyond the model’s knowledge cutoff, and when automatic or manual search occurs.

Learn how the deep research tool functions as a research assistant in ChatGPT, generating search terms, aggregating and evaluating sources, and planning extended web research.

Discover how ChatGPT's image generation works: describe an image and let AI generate it, often without enabling the create image tool, with optional explicit mode if results misaligned.

Enable canvas mode to view and edit long AI responses in a separate window and annotate or refine specific parts of text. Select the text and ask ChatGPT to edit.

See how ChatGPT's code generation and sandboxed Python execution create and run charts, and generate QR codes that link to webpages.

Tool availability varies by plan and model, with changing tool names and limited access, such as image creation not available on ChatGPT 4.5.

Discover how to customize ChatGPT for all future chats by setting personal details, traits, what it should know, and tool controls like code execution and image generation.

Explore how to customize ChatGPT through appearance, memory, language, chat management, notifications, and privacy controls, plus data export and delete options.

Create chat projects to group chats, files, and instructions in one workspace. Upload PDFs and CSVs for project-wide analysis and extract insights across shared documents.

Explore the ChatGPT tasks feature to create recurring tasks for daily ai news summaries, with Try Your Own options, and model-specific scheduling.

Explore custom GPTs, pre-configured chats with built-in instructions and documents loaded to enhance knowledge, via the GPT marketplace, including actions with third-party APIs.

Create a custom GPT by defining base instructions and configuring name, description, and tools to enable reusable, shareable templates with optional API actions.

Learn to build a more realistic custom GPT focused on text summarization, configured to extract key concepts and present concise bullet-point summaries for any input.

Take a concise tour of ChatGPT's core features, including managing conversations, prompts, and settings, and preview image generation, data analytics, and prompt engineering.

Survey alternative ai chatbots beyond ChatGPT, including Google Gemini, Anthropic Claude, xAI's Grok, and Microsoft Copilot. Compare their features and usage plans with ChatGPT to understand options and differences.

Discover Google Gemini as Google's answer to ChatGPT, featuring advanced models, Gemini Advanced, longer context, and deep research for planning and web results.

Gemini uses web search by default to improve responses and pull sources from websites. Use the double-check feature to verify results and see which parts are source-supported.

Explore how Google Gemini can be accessed indirectly via google.com to gain an ai overview without an account, though it's not a chat, with responses backed by web search results.

Explore Anthropic Claude, including free access, pricing, chat history, customizable styles, per-prompt writing styles, file uploads, code-focused Sonnet models, and model switching for coding tasks.

Explore Grok, X.ai's integrated AI chatbot in X, with web-grounded responses, file uploads, deep research mode, image editing, and explain post features for researching and aggregating X content.

Explore how Microsoft Copilot operates across apps and the Edge browser, including free vs paid plans, chat modes, file attachments, and voice mode with web-aware insights.

Explore the limitations of proprietary models and learn how open alternatives like Gemma, LLaMA, and DeepSeek can run locally via quantization and LM Studio.

Compare the costs, availability, privacy, and usage limits of using ChatGPT and other providers versus running open-source large language models locally, and consider when open-source options like DeepSeek might fit.

Discover open LLMs you can run locally and for free, compare open source options on Hugging Face, and learn how benchmarks and licenses guide selection.

Explore open-source models like Meta's LLaMa and Google's Gemma, and learn how quantization reduces memory needs to run them locally on GPUs or CPUs.

Discover how to run open-source LLMs locally using LM Studio, Ollama, or Llama.cpp, with quantization to 4-bit, 8-bit options, and programmatic access via Hugging Face Transformers.

Explore open source models like the Gemma, with a 128,000-token context length and offline use, to summarize PDFs and extract data locally with LM Studio.

Host open source models on your own or rented servers using tools like Ollama, LM Studio, or LLaMA.cpp, expose via API, and weigh self-hosting versus managed Bedrock.

Master prompt engineering concepts and techniques to ensure good results when working with generative AI. Explore examples, best practices, and recommendations relevant to anyone using generative AI.

Define prompt engineering and explain why prompts impact ai responses. Show how adding specific context and extra instructions improves results in chat interfaces and other message-based interactions.

Explore how prompt engineering uses customization and general instructions to shape ChatGPT's behavior; customize responses to be brief or more chatty, a feature not available on all chatbots.

Explore how prompt engineering controls both the output content and the output format, shaping tone, length, and structures like Markdown, JSON, or plain text.

Craft effective prompts by adding clear instructions and helpful context, including roles and constraints. Understand how token prediction and context windows influence the AI's responses for task-specific results.

Master prompt engineering to treat generative ai output as a starting point, then refine via follow-up prompts and human edits to blend efficiency with expertise in chat workflows.

Practice prompt engineering techniques and apply tips to draft prompts for a product announcement, a Python file deletion script, and a feedback email by the end of the week.

Craft prompts with role and channel context to generate a concise product announcement for an AI-powered website generator, targeting non-tech-savvy users on LinkedIn and X.

Solve the second task by generating a Python script that searches and deletes PNG and JPEG files in a folder. The lecture discusses prompt engineering and delivering code-only results.

Practice crafting an urgent yet polite email draft to request feedback on a submitted prototype by week's end, using context, tone, and clear instructions.

Learn specific prompt engineering techniques to craft good prompts by combining instructions with helpful context and examples. Some techniques contradict, so you choose wisely when needed.

Explore zero-shot and few-shot prompting techniques, using examples to steer tone, style, and content while avoiding content leakage and managing context window limits.

Compare few-shot prompting and fine-tuning, explaining when to include examples, how fine-tuning creates a task-specific model, and why it may be expensive or reduce performance on other tasks.

Learn how few-shot prompting uses a single example and delimiters to steer AI output, shaping tasks like word-by-word translation instead of a full sentence.

Use delimiters to structure prompts and boost prompt engineering for few-shot prompting. Mark content with XML tags or similar markers to clearly enclose captions and improve AI summary quality.

Include relevant data and examples in prompts to guide AI behavior. Use web search and upload files to summarize articles, while monitoring context window and token usage to avoid hallucination.

Explore chain-of-thought prompting to guide AI through step-by-step reasoning for data analysis and logical problems, clarifying how token prediction differs from calculation and improves accuracy.

Learn how reasoning models include built-in chain-of-thought prompting, and why you should avoid adding step-by-step prompts when using them for logical and math problems.

Learn to use ask-before-answer prompting to have ai ask clarifying follow-up questions before delivering a final result, improving response quality.

Explore persona prompting by defining role descriptions, refining the AI persona, and detailing target audiences to improve responses. Learn to tailor prompts with context and problem guidance for optimal results.

Use self-reflective prompting to improve AI outputs by analyzing strengths and weaknesses, generating revised versions, and applying follow-up prompts within the context window.

Explore negative prompting to exclude unwanted aspects from AI responses, such as ads-related advantages when comparing generative AI with search engines. Learn how this technique can pair with prompt strategies.

Master controlling the AI output format to produce csv, markdown, or lists with data-only outputs for reliable integration in downstream workflows.

Generate prompts with generative AI to kick-start your prompting process, then customize them with your own knowledge. Ask the model for prompting tips to craft the best possible prompts.

Explore essential prompt engineering techniques that apply across AI providers and models, and learn when to consult provider-specific guides from Anthropic and OpenAI for better results.

Explore essential prompt engineering techniques, tips, and recommendations, and learn to combine methods and experiment with prompts to achieve reliable results.

Practice advanced prompting techniques through hands-on tasks, building a modern CV website prompt, drafting a newsletter email with ask-before-answer prompting technique, and generating multi-language code that shifts text left.

Use one-shot prompting to give ChatGPT a fixed template and content, generating HTML and CSS for a modern minimal website. Define the exact structure and content, including positions and companies.

Explore the ask-before-answer prompting technique to craft engaging newsletters, including role assignment, goals, context, and how to gather necessary information before drafting blog-promoting emails.

Practice advanced prompt engineering with ChatGPT by using few-shot prompts, assigning Python and JavaScript developer perspectives, and generating runnable Python and JavaScript implementations.

Learn how contextual prompting adds cultural, historical, and situational context to prompts, balancing a clear goal with extra information to improve responses in evolving topics like energy and technology.

Harness emotional prompting to have ChatGPT consider user sentiment when drafting responses, such as calming a frustrated customer email, and reduce follow-up prompts.

Master laddering prompting to break complex problems into smaller prompts, build modular blocks, and compose a Node.js REST API step by step with dummy code.

Define a clear goal with contextual information, then apply prompting techniques like laddering, emotional, contextual, and perspective prompting; use role setting and one shot prompts to extract strategies from ChatGPT.

Collaborate with ChatGPT by uncovering missing information to improve outputs, using ask-before-answer prompts and strategic follow-ups to refine responses.

Explore self-evaluative prompting by asking ChatGPT to criticize, evaluate, and improve its own responses, revealing when self-critique enhances results and when it may not.

Learn how to use ChatGPT to split complex tasks into manageable steps using laddering prompting, guiding the model to research keywords and outline a multi-step plan.

Explore chain-of-thought prompting by forcing ChatGPT to outline its reasoning, improving results in problem solving and metric selection through step-by-step derivation and self-reflection.

Learn how ChatGPT self-critiques and improves, reveals what extra information boosts results, and uses prompting strategies to craft efficient prompts, including perspective prompting and example-output prompts.

Master crafting ChatGPT prompts by structuring concise prompts with clear context and audience, applying zero, one, or few-shot techniques, and using inspiration from prompts found online as a starting point.

Explore prompts published by others to craft your own, and assess super prompts like CAN (code anything now) and DAN, noting they're over-engineered and less powerful than general prompting techniques.

Learn to craft prompts by pairing goals with context and constraints, define audience and format, and apply zero-, one-, few-shot prompting, ask-before-answer prompting, and perspective prompting.

Explore generative AI beyond chatbots, including image and video generation, and learn how to select models and control image outputs and formats.

Explore AI image generation options from chatbots like ChatGPT and OpenAI API to dedicated models like Midjourney and Stable Diffusion, including Adobe Photoshop integrations and use cases.

Learn how to generate images with ChatGPT by using prompts or the create image button, explore model differences, image styles, and potential pricing or availability changes.

Explore alternatives to ChatGPT for image generation, compare Google Gemini and Grok, and assess quality over time to decide which provider you prefer.

Explore AI image generation beyond chatbots with Midjourney and Photoshop; use the Midjourney Discord bot to generate four variations and upscale your favorite.

Explore integrated ai image generation in Photoshop as a practical example, crafting prompts, viewing variations, and refining results with ChatGPT in the complete guide to generative ai.

Learn to craft specific prompts for AI image generation with ChatGPT, including role assignment, style, pose, and context, to guide the model toward the desired image.

Improve image generation with AI chatbots by applying prompt engineering, provide helpful context, and use ask-before-answer prompting to guide follow-up questions and image results.

Explore the limitations of AI-generated images, including hallucinations and factual errors, and learn to use AI tools with caution rather than blind trust.

Use AI image generation tools, including ChatGPT, to create any kind of illustration, from logos to icons, beyond photorealism. Describe details and color schemes in your prompts for precise results.

Refine images over time by describing changes in chat messages with ChatGPT, keeping the image in the chat history and fine-tuning with follow-up prompts.

Explore how to control image aspect ratios in AI image generation, using prompts with ratios like 16:10, and combine aspect ratio with other image details in ChatGPT.

Explore how ChatGPT’s image generation can include legible text in images, such as neon signs, and note that long text may still contain errors, requiring verification.

Explore how ChatGPT and AI models analyze uploaded images, accurately describe what they see, and how to include images in prompts to guide image generation and interpretation.

Upload images as prompts and generate new images based on the uploaded base, placing you in a lake scene and adjusting details like a blue shirt.

Upload a sketch to describe and generate website code matching the image, then use canvas mode to view, edit, and preview a header with left posts and right news.

Learn to generate and edit videos with Sora in ChatGPT, switch from image to video, set 16:9 and 720p, and turn images into moving video via prompts and storyboard.

Explore AI video generation options, including Sora, Kling.AI, and Veo 2, and learn Veo 2's current 720p landscape, eight-second limits, and common model mistakes, with advice to test multiple models.

Explore practical uses of chatgpt for daily tasks, including researching, summarizing and translating, proofreading for grammar, improving text, and building simple Python-based utility tools with no coding required.

Learn to conduct efficient research with ChatGPT: start with generic prompts, refine with constraints for business or personal development, and extract titles and publication details while verifying results.

Learn to use ChatGPT to summarize text quickly, with summaries that preserve core topics and subheaders, and paste articles when needed; note that ChatGPT cannot browse websites by default.

Explore translations and tone adjustments with ChatGPT, translating across languages and shaping friendly, respectful emails and feedback, including paragraph-by-paragraph outputs and tailored formats.

Use ChatGPT to proofread and enhance text, correcting spelling and grammar and rephrasing to a more native English style for emails and other documents.

Learn how ChatGPT helps you create Python utility scripts to automate daily tasks. The lecture shows installing Python, creating rename_files.py, and running the script to convert uppercase filenames to lowercase.

Discover how generative AI, including ChatGPT, can generate content and email drafts while supporting research and data analysis. Learn practical ways to use these AI models for research tasks.

Learn to use AI-assisted web search for research tasks, verify results to avoid hallucinations, and treat AI as a research assistant rather than a replacement.

Learn how to perform deep research with AI, conducting multiple site searches, planning steps, and using Python code to analyze market caps across dates, while evaluating sources and results.

Leverage ChatGPT and AI tools to analyze uploaded data, generate Python code, and create summary statistics and a sales by month chart, then verify results.

Explore how ChatGPT generates interactive charts for data analysis, allowing you to tweak colors, switch to interactive mode, and hover to see exact figures like revenue by products.

Analyze unstructured data with ChatGPT and large language models to extract key metrics from documents like pdf reports and generate tables or charts using optional code execution tools.

Learn how chatbots like ChatGPT extract text and numbers from scanned documents and photos, handle unstructured data with multimodal models, and the importance of double-checking results to avoid hallucinations.

Explore canvas mode in ChatGPT and other AI services for writing and editing code within an integrated canvas. Experience code selection, in-browser previews, and collaborative editing to refine AI-assisted projects.

Discover how ChatGPT's code interpreter enables generating and executing Python code in a restricted, isolated environment for data analysis, image editing, and data transformation, with clear limits.

Enable the advanced data analysis feature’s code interpreter in GPT-4 to generate and run Python code, view the work, and maintain a shared execution environment for iterative calculations.

Upload files to ChatGPT with Code Interpreter to capitalize text in an isolated environment within a 100 megabyte limit. See how it handles acronyms like NLP, AI, and GANs.

Explore how ChatGPT can generate new files and expose download links. Save the generated text locally as a .txt file.

Explore prompt engineering with code interpreter enabled to craft prompts, supply investment data, and generate verifiable results; learn how follow-ups fix formatting and incorrect code.

Explore prompt engineering with code interpreter to force ChatGPT to generate and execute Python code, producing hex color schemes from a base color.

Master a prompt engineering technique to disable the Code Interpreter for specific parts of the task, guiding ChatGPT to open a text file with Python and summarize content itself.

Explore using ChatGPT with Code Interpreter to extract data from files, such as PDFs, by uploading documents and requesting details like address, invoice number, and amount billed.

Generate dummy data using ChatGPT and the code interpreter to create, execute Python code, and produce a downloadable sales dataset.

Explore a simplified data analytics workflow: data preparation, data analysis, and data visualization. Load and clean data, apply descriptive statistics, then create charts to visualize results.

Prepare and clean a sales dataset using python and pandas through the code interpreter. Fix separators, handle missing values, convert data types, and produce an updated csv for analysis.

Explore descriptive statistics and data visualization using the code interpreter, including mean, standard deviation, and weekday averages, plus bar and line charts of monthly sales.

Explore analyzing and transforming images with a code interpreter to extract metadata and text, generate new images, and edit by cropping to preserve aspect ratios without distortion.

Extract a top color palette from an image, display colors as squares with hex codes, and edit images using code interpreter. Save edits as grayscale or JPEG for download.

Explore how OCR enables the code interpreter to extract text from images, using a dummy seafood menu to show extraction, limitations, and refining steps as a first-use case.

Explore how ChatGPT drives content creation from blog posts and articles to video scripts, social media, and CVs. Generate sample data in JSON, Markdown, or ZSV.

learn to generate engaging articles with chatgpt by crafting a simple prompt, then a structured one with constraints, audience, seo keywords, and image descriptions for iPhone history.

Develop an article by first listing seo keywords, creating an outline, and refining prompts, then write within 1,200 words on iPhone topics like history, design, and ecosystem.

Use a single prompt to generate an SEO article with outlines, sections, subheaders, images, and iPhone development focus. Define a professional blogger role, target audience, and SEO keywords.

Learn to generate at least five creative article titles and a concise executive summary using prompts, starting from an initial article and refining beyond the first result.

Brainstorm content with ChatGPT across tutorials, opinions, articles, and deep dives on tech, fitness, beauty, and social topics, then refine and expand ideas using the base ChatGPT.

Convert a full article into a tailored YouTube video script with ChatGPT, enforcing an 8–10 minute, fast-paced format plus footage, background music, time-stamped sections, a video name, and thumbnail.

Learn how copywriting uses targeted, engaging text to promote products and services across website content, social media, and email campaigns, enhanced by ChatGPT.

Craft landing page copy by assigning a ChatGPT copywriter role, defining a target audience of 14–21, and using structured prompts with bullets, emojis, and background image guidance.

Boost your social media accounts with ChatGPT to craft motivational posts for Instagram and Twitter, refine prompts, and generate image descriptions.

Learn how to prepare engaging podcast interviews with chatgpt by shaping tone and icebreakers, crafting project- and movie-related questions, and iterating prompts to refine the interview.

Discover how ChatGPT and generative AI streamline job-related writing with practical examples of standout CVs, compelling cover letters, employer drafts for job offers, contracts, and work-related emails.

Create structured CVs and personalized cover letters using advanced prompts in ChatGPT, detailing education, work experience, and achievements to tailor your application and stand out.

Use ChatGPT to craft tailored job offers for tech roles, emphasizing learning, collaboration, and teamwork, and to generate working contracts from the offer, with a creative, youth-friendly headline.

Discover how to craft a personalized resignation email with ChatGPT, sharing details and a career-goal aligned reason for quitting, creating a specific, heartfelt farewell to teammates.

Generate and customize dummy data with ChatGPT using JSON, CSV, table, and Markdown formats for projects from recipes to city datasets; leverage prompts to tailor outputs and copy-paste-ready results.