TLDR;
This YouTube video provides a comprehensive guide to using AI tools for various tasks, including prompt engineering, productivity, website building, image generation, audio and video creation. It introduces key concepts like prompt engineering using the CREATE technique, and PROFILE formula. The video also explores various AI tools, ranking them based on their usefulness and providing live examples of how to use them effectively.
- Prompt engineering using the CREATE technique
- Productivity tools such as Napkin AI, Gamma AI, Chronicle AI, Jark AI, and Notebook LM
- Website building using Framer AI and Lovable AI
- Image generation using various AI models and the PICTURE technique
- Audio and video creation using AI tools like Suno AI, Eleven Labs, Speechma, Whisper Flow, Kling AI, and Hugging Face
Introduction [0:00]
The video introduces eight chapters covering prompt engineering, productivity tools, website building, image generation, audio and video creation, honourable mentions, and AI chatbots. The presenter states that live examples will be provided where possible to enhance understanding.
AI Chatbots and Prompt Engineering [1:03]
The video introduces AI chatbots like ChatGPT, Gemini, and Perplexity, explaining that they allow users to communicate through chat. The messages written to these bots are called prompts. The quality of the responses from these AI chatbots depends on the quality of the prompts. The skill of writing effective prompts is known as prompt engineering, which is not an engineering field but a way to get useful answers from AI.
The CREATE Prompt Technique [2:02]
The video introduces the CREATE technique to improve prompts. CREATE stands for:
- Character: Define the AI's persona (e.g., "act as a world-class travel planner").
- Request: Clearly state the request (e.g., "plan a seven-day trip to Goa").
- Example: Provide examples or inspiration (e.g., "take inspiration from top family vacation books").
- Adjustment: Include specific adjustments (e.g., "focus on healthy dining and fitness activities").
- Type of Output: Specify the desired output format (e.g., "present the itinerary in a detailed table").
- Extra Guidance: Add any extra instructions (e.g., "explain like I'm five years old").
The video demonstrates how using the CREATE formula results in more detailed and useful responses from AI.
Live Example of CREATE Technique [6:12]
The presenter demonstrates the CREATE technique live using AI Fiesta, an AI chatbot platform. The example involves providing PR advice to Arjun Kapoor, who is facing trolling on the internet. The presenter defines the character as a PR expert, specifies the request, provides examples, and asks for the output in a table format. The AI generates a detailed, day-by-day PR strategy. The presenter also highlights AI Fiesta's feature to generate a consensus from multiple AI models.
Project Feature for Efficiency [12:03]
The video introduces the project feature available in AI chatbots like ChatGPT, Gemini, and AI Fiesta. This feature speeds up work by saving prompts and information. The presenter demonstrates creating a project for a PR expert, saving the character definition and initial prompt. This allows for quicker, one-line questions in the future, as the AI already has the background knowledge.
Profile Formula for Personalised AI [14:51]
The video introduces the PROFILE formula to make AI sound more like the user. By providing AI with personal writing styles and preferences, the AI can generate responses that match the user's tone. The presenter demonstrates creating an email writer project, inputting their writing style, and generating emails that reflect their personal style.
AI Fiesta's Expert Team [20:02]
AI Fiesta offers an expert team feature, providing pre-built projects like a personal doctor, finance advisor, career coach, and fitness coach. These avatars ask specific questions to gather detailed information, creating a personalised AI experience. The presenter demonstrates the personal doctor feature, explaining how it uses health information to provide tailored advice.
Productivity Tools: Napkin AI [25:31]
The video transitions to productivity tools, starting with Napkin AI, which visualises text. Users can copy-paste text or generate it via AI, and Napkin AI creates visualisations such as diagrams and flowcharts. The presenter shares a personal use case, visualising data from a script about the Maha Kumbh Mela. Napkin AI is useful for presentations, offering customisation options and a generous free plan.
Gamma AI for Presentations [28:45]
Gamma AI is introduced as a tool for creating PowerPoint presentations using AI. It offers three options: generating a presentation from a one-line prompt, pasting rough notes, or uploading a PDF. The free version allows up to 10 slides. The presenter demonstrates creating a presentation on the Mughal history of India, customising the outline, style, and image sources. Gamma AI also functions as a PowerPoint, allowing users to edit and export presentations.
Chronicle AI and Jark AI [37:00]
Chronicle AI is presented as a competitor to Gamma AI, offering unique designs. Jark AI, a Chinese AI, allows users to create presentations through chat, generating code that turns into a presentation. Jark AI is free to use and offers professional-looking designs.
Notebook LM for Students [38:39]
Notebook LM, a tool by Google, allows users to restrict the knowledge base used by AI to specific sources. Users can create notebooks with files, YouTube videos, and text, and then chat with the AI using only those sources. The presenter demonstrates creating a notebook on evolution, using their own video, a textbook chapter, and a book by Charles Darwin. Notebook LM also offers a studio function to create mind maps, infographics, quizzes, audio podcasts, and reports.
Website Building with Framer AI and Lovable AI [46:53]
The video moves to website building, initially mentioning Framer AI, which generates websites from prompts. However, the presenter shifts focus to Lovable AI, a more accurate and user-friendly tool. Lovable AI allows users to create websites by providing a prompt and engaging in a chat-based interface. The presenter demonstrates creating a CV website using Lovable AI, inputting information and customising the design.
Live Demonstration of Lovable AI [49:14]
The presenter provides a live demonstration of Lovable AI, creating a CV website. They input information, choose a creative and colourful visual style, and add animated stats. Lovable AI generates a plan for the website, which the presenter approves. The presenter also showcases a website they created for converting scripts to audio using the Eleven Labs API.
Customising Lovable AI Websites [59:09]
The presenter continues customising the CV website, adding a profile photo. They explain how to add elements and modify the website by simply typing instructions. Lovable AI offers five free credits per day, allowing users to test the tool. The presenter demonstrates publishing the website, making it live on the internet.
Advanced Website Customisation and Domain Connection [1:02:49]
The presenter explains how to connect a custom domain to a Lovable AI website using GoDaddy. They also demonstrate improving an existing website using Lovable AI. For advanced animations, the presenter recommends 21st.def, a website with free templates that can be integrated into Lovable AI. The presenter showcases adding a robot animation and button animations to a website.
AI Chatbots for Simple Websites [1:12:42]
The video explains how to create simple websites using AI chatbots like ChatGPT, Gemini, or AI Fiesta. The presenter demonstrates creating an HTML website wishing Happy Diwali using AI Fiesta. The AI generates the code, which can be downloaded and opened in a browser. For publishing, the presenter suggests using Netlify.
Image Generation: The PICTURE Technique [1:18:36]
The video transitions to AI image generation, introducing the PICTURE technique to improve image prompts. PICTURE stands for:
- Photographic Style: Choose a style (e.g., cinematic, vintage).
- Imagery/Scene: Describe the scene and subject.
- Camera Angle & Composition: Specify the camera angle (e.g., close-up, low angle).
- Timing & Lighting: Define the lighting (e.g., sunset, harsh noon light).
- Use Film/Effect: Add effects (e.g., lens flare, motion blur).
- Render Style: Choose a render style (e.g., hyper-realistic, comic book).
- Exact Details: Specify skin texture, hair detail, etc.
AI Image Generation Models [1:20:48]
The video ranks various AI image generation models:
- GPT 1.5 Image (B Tier): Easy to use but has high censorship and struggles with real images.
- C Dream 4.5 (A Tier): A top Chinese tool with high-quality images but difficult to access and has Chinese censorship.
- Ediagram (B Tier): Offers character features and high-quality outputs within its templates but has limited free usage.
- Meta AI (B Tier): Free to use and uses the Midjourney AI model but has watermarks, is unreliable, and collects data.
- Grok AI (B Tier): Generates unlimited images but has low resolution and collects data.
- Nano Banana Pro (S Tier): The best AI image-generating model with hyper-realistic images and editing capabilities but is paid and adds watermarks in the Google app.
AI Fiesta's Image Studio [1:34:20]
AI Fiesta offers an Image Studio, allowing users to use Nano Banana Pro, GPT's latest image generation, and C Dream 4.5 simultaneously. The presenter demonstrates generating images of a man riding a motorcycle in the forest, showcasing the different outputs from each model. AI Fiesta also offers an integrated image editor for further customisation.
Identifying AI Images and Image Upscaling [1:41:02]
The video discusses identifying AI-generated images, noting that it is increasingly difficult. Google's SynthID can only identify images made with Google's tools. The video also mentions Magnific AI, a tool for upscaling images, but notes that AI Fiesta also offers an upscaling feature.
Video Generation: AI Models and Techniques [1:43:24]
The video transitions to video generation, covering three main techniques: text-to-video, image-to-video, and video-to-video. The video tests and ranks various AI video generators:
- Sora 2 AI (C Tier): Slow and generates sloppy videos with strict policies.
- Halu AI (B Tier): Generates videos from images but lacks audio and has limited resolutions.
- See Dance 1.5 Pro (A Tier): Adds audio and follows prompt instructions well but is limited to 720p.
- Vain 2.6 (B Tier): Adds lip sync to AI videos but has hit-and-miss physics.
- Kling AI (S Tier): The gold standard for AI video generation with excellent physics and prompt following.
- Meta AI (C Tier): Generates videos without sound and has poor quality.
- Grok AI (C Tier): Generates videos with poor quality.
- Google Vevo 3.1 (A Tier): Competes with Kling AI but is slightly inferior and has limited access.
Kling AI and Lip Sync AI [1:50:32]
The presenter highlights Kling AI as the best video generation tool, showcasing its motion brush and lip-syncing capabilities. They demonstrate translating a video into English with accurate lip movements. The presenter also recommends Lip Sync 2 Pro as the best AI for lip-syncing.
Audio Generation: Suno AI and Eleven Labs [1:59:03]
The video moves to audio generation, starting with Suno AI for creating songs. The presenter demonstrates generating a Bollywood love song and a remix of "Mere Sapno Ki Rani." Suno AI offers a generous free version and has excellent copyright policies. For generating voices, the presenter recommends Eleven Labs, a text-to-speech AI platform.
Speechma and Whisper Flow [2:05:12]
Speechma is introduced as a free Chinese alternative to Eleven Labs for generating short texts. For speech-to-text, the presenter recommends Whisper Flow, an AI software with an iOS and Android app. The presenter also highlights AI Fiesta's transcription tool, which converts speech to text.
Honourable Mentions and Combining AI Tools [2:08:14]
The video mentions Sam to for rotoscoping and Hugging Face for VFX generation. The presenter showcases a रामायण movie teaser created by combining various AI tools. The presenter outlines their workflow for creating AI videos, including generating scripts, creating audio, generating images, and creating videos.
