Gen AI and The Learning Revolution
“Imagine your AI is a big recipe book. Instead of saying, ‘Give me something tasty,’ you’re guiding it: ‘How about a dessert with chocolate?’ Prompt engineering is tweaking your question to get the yummiest answer.”
- ChatGPT Prompt Genius - Free & Open Source
- Sider - ChatGPT Sidebar with GPT-4 - Enhanced when signing in
Spaced learning
AI Models
“Imagine your brain as a sponge that soaks up information. Now, picture a machine learning model as a mini, digital sponge. Instead of soaking up memories of that time you fell off your bike or your favorite recipe, it absorbs patterns from loads of data. And just like you might recall how to ride your bike, the digital sponge ‘remembers’ these patterns to make predictions or decisions. That ‘memory’ it forms after soaking up all that data? That’s a trained model.”
AI Generators
AI generators are software programs that use artificial intelligence, particularly machine learning algorithms, to generate content or data. They can be used in a variety of fields, including text, image, music, and more. These generators are typically trained on large datasets and can produce new, unique content based on the patterns they’ve learned.
Here are some examples of AI generators:
-
Text Generators: These are perhaps the most common type of AI generators. They are used to create written content. For example, OpenAI’s GPT-3 (which powers this assistant) is a text generator that can write essays, answer questions, translate languages, and more.
-
Image Generators: These AI tools can create unique images based on certain parameters. For instance, DeepArt and DeepDream use AI to transform images into artistic pieces. Another example is This Person Does Not Exist, which uses Generative Adversarial Networks (GANs) to create images of people who don’t actually exist.
- DALL·E is an AI image generator developed by OpenAI. It’s essentially a 12-billion parameter version of GPT-3 trained to generate images from text descriptions. The system uses a dataset of text-image pairs to learn how to create images based on textual input. It has a diverse set of capabilities, including creating anthropomorphized versions of animals and objects, combining unrelated concepts in plausible ways, rendering text, and applying transformations to existing images. DALL·E can generate images from scratch or regenerate any rectangular region of an existing image that extends to the bottom-right corner, in a way that is consistent with the text prompt.
- Midjourney:
- Stable Difussion: Open-Source you can run it locally or using cloud computing power
-
Music Generators: AI can also be used to create music. Tools like OpenAI’s MuseNet can generate original pieces of music in a variety of styles and genres.
-
Code Generators: AI can even generate code. GitHub’s Copilot, powered by OpenAI Codex, is an AI tool that suggests lines or blocks of code as developers type.
-
Chatbot Generators: These AI tools can generate human-like text based on user input. They are often used for customer service, sales, and marketing. Examples include Mitsuku and Replika.
-
Data Generators: AI can generate synthetic data that can be used for testing or training other AI models. Tools like Tonic and Hazy can generate realistic, but not real, data for these purposes.
Prompt Engineering
-
Think step by step.
-
Explain [topic] starting by simple and easy terms that any beginner can understand.Then level up and continue with a paragraph to intermediate level, then advance. Try to not repeat the same sentences.
-
Compare [options 1,2,3] starting by simple and easy terms that any beginner can understand.Then level up and continue with a paragraph to intermediate level, then advance. Try to not repeat the same sentences and include pros and cons of the options
-
Explain the topic of sales and branding for data analytics in 4/5 paragraphs starting by simple and easy terms that any beginner can understand. Then level up and continue with a paragraph to intermediate level, then advance. Try to not repeat the same sentences.
-
Content Creation:
- Write me a seo tittle better than: Crypto 101 - Understanding the basics
- Tell me a 100 char description for that blog post about xxx for data analytics and also a 50 words summary (seo friendly)
- Write an engaging introduction paragraph for a blog post about : how to build your brand for data analytics, also a cool title for seo
ChatGPT Plugins
-
voxscript: extract summary from yt video
-
Video Insights: it can summarize youtube videos
-
Code Interpreter: my favourite
-
Web Pilot: it connects to the internet
-
Website Performance:
-
Wolfram: to do cool Math stuff
PPTs with GenAI
PPT Generation
Approach 1
- The Prompt:
- Act like a XYZ profesional. The topic is XYZ. Create a detailed outline with bullet points based on the topic.
- I am creating a ppt. The topic is ChatGPT plugins. Create a detailed outline with bullet points based on the topic with the most popular options that have been released so far.
- Convert this to HTML
- Copy paste and save as html
- Open the html in word and change the file format to .docx
- Upload it to onedrive and open it in Office365
- Convert it now to Powerpoint (Home -> Export)
Approach 2 - With VBA
VBA can Automate Microsoft Apps tasks, and PowerPoint is just one of them.
- Ask the LLM to generate PPT Slides about your Topic and to write the VBA code that would generate the slides.
- Then go to PowerPoint - File -> Options -> Customize Ribbon -> Tick the Developer Option
- Click the Developer icon -> VBA - (ALT+F11)
- Insert a Module and paste the VBA Code
FAQ
What it is the Ziegarnik Effect?
The psychological observation that incomplete tasks are more easily remembered than completed ones. This effect can be leveraged by students to enhance study and task management.
By starting assignments or study sessions and leaving them intentionally incomplete, students may find an increased motivation to complete these tasks due to the mental tension created by their incompletion.
This effect helps in maintaining better recall and focus on the task at hand, but it’s important to balance it to avoid feeling overwhelmed by too many unfinished tasks.
That’s actually the point on keeping a low To-Do in Kanban Boards.
For effective learning and productivity, understanding and utilizing the Zeigarnik Effect can be quite beneficial, as it can help in breaking down large tasks into smaller, more manageable parts, thereby enhancing overall academic performance.
What is Spaced Learning?
Spaced learning, or spaced repetition, is a learning technique where information is reviewed at increasing intervals over time.
This method leverages the psychological spacing effect, where learning is spread out over time, rather than crammed in a single session. It’s based on the idea that our brains better retain information when exposure to it is spaced out rather than concentrated.
Spaced learning optimizes the timing of learning sessions for better memory retention, the Zeigarnik Effect can enhance the focus and engagement within those sessions by creating a psychological urge to return to and complete the learning tasks.
Diagrams with GenAI
There is an awsome F/OSS tool - Mermaid, which we can combine with LLM models to make powerful diagrams.
Timeline
timeline
title A great Timeline Chart
Section1 : Task A
Section2 : Task B : Task C : Task D
Section3 : Task E : Task F
Gantt
gantt
title A Great Gantt - Upcoming Tweaks
dateFormat YYYY-MM-DD
%% excludes weekends 2023-12-24, 2023-12-25, 2023-12-31, 2024-01-01
section Section 1
Task 1 :od1, 2023-12-10, 30d
Task 2 :after od1, 25d
section Section 2
Task 1 :ds1, 2024-01-01, 15d
Task 2 :after ds1, 15d
Task 3 :after pc1, 10d
section Section 3
Task 1 :ex1, 2023-12-18, 5d
Task 2 :after ex1, 10d
Task 3 :after dpc1, 15d
Task 4 :after pd1, 25d