What's Inside This Week's Newsletter:
- This Business Idea Still Has Low Competition in 2024!
-
17 Tools & Resources for REAL Prompt Engineers​
- Watch me Build an AI Video Summarizer in 57 seconds!
- Let's Learn About AI Agents here!
- My Latest Research & Updates.
- Quote of the Week 💬
This Business Idea Still Has Low Competition in 2024!
I’m sharing with you today a business idea that I implemented myself, and it not only helped me establish a new income stream but also led to significant growth in sales for my main products.
I’m talking about building "Data-Driven" tools. You can check my data library [here] to get an idea of what I’m talking about.
In short, it’s all about building a simple tool that gives people access to important data.
The great thing about this strategy is that once you know what type of data you want to collect, the rest is easy. It requires no coding and still has low competition.
I created a full, free video guide with more examples and ideas, and I showed how to implement practically it in case you want to try it yourself!
​
17 Tools & Resources for REAL Prompt Engineers
A couple of days ago, I shared on Twitter (X) a list of tools, resources, and libraries that I use almost daily while creating LLM-powered projects, whether for research or building new tools.
Since this is my new Twitter account, you probably didn't see it.
​
Watch me Build an AI Video Summarizer in 57 Seconds!
A couple of days ago, I published a 57-second video showing how I built an AI-powered YouTube Video Summarizer using SimplerLLM.
'If you don't want to watch the full 57-second video😅, here is what I did:
- Imported SimplerLLM.
- Used the built-in function
load_content to load the video script.
- Asked the LLM to generate a summary with a simple prompt like this:
"Generate a concise summary in bullet points format for the following video script: {video_script}."
Here is the code:
And in case you want to see it in action in 57 seconds, here is the video:
​​
Let's Learn About AI Agents here!
Yes, today and from now on, I will be crafting this section and explaining AI Agents from scratch. The goal is to learn something new about Agents every week in just 1 minute! What do you think? Let's try it, and let me know your thoughts.
What are AI Agents?
An Autonomous AI Agent integrates a large language model (LLM) with external functions and enhanced prompting mechanisms.
This system, known as an "AI Agent," allows you to mimic humans in the way they achieve tasks.
For example, if I ask you now, "What is the speed of your internet connection?"
The first thing you would think about is using a tool to test the speed and then provide the answer.
However, an LLM would say, "I don't have access to such data!"
but, By giving it a tool to use and prompting it to think before answering, the LLM will know if it needs a tool before responding, thus mimicking human behavior.
Agents are the future, so it's important to get into this as soon as possible!
​
My Latest Research & Updates.
In the last couple of months, I have been working on fine-tuning GPT-2 (500 MB) for RAG systems. Two days ago, I got some interesting new results:
✅ The new GPT-2 outperformed GPT-3.5 in 167 out of 1000 queries!
✅ The new GPT-2 beat the base model 977 times!
The evaluation was done with GPT-4, which provided a score for each model in a dataset of 1000 queries.
There’s real potential in GPT-2 🚀
What do you think?
​
Quote of the Week 💬
"Learning a new skill is an investment in yourself that pays dividends for a lifetime. Start today and see how far you can go"
See you next week with new pulses!
Hasan
