Typically, with a search engine you give a list of keywords in an input box and the engine scours through the latest pre-crawled list of websites relevant to that list of keywords.
The issue with a search engine, in regards to learning, is that it is optimized for keywords and finding specific phrases on a website and not specifically for answering your question, but for finding the most relevant websites.
Large Language Models(LLMs), however, are built to specifically answer your question as if you were talking to a human. Some of the best LLMs give such specific, relevant, advice and at such length, it is invaluable to your learning.
The reason this happens is due to a mathematical algorithm which emulates human thought, the neural network…
When to Ask an LLM Questions
Almost always, while I am coding, I use Chat-GPT to enhance my learning speed. It is invaluable in how it can answer questions about my framework or language.
Recently, I was learning Docker and had been asking Chat-GPT questions as I went through a monolithic course on the subject. There were many instances where the instructor of the course did not explain or clarify a detail in enough depth as I would have liked.
At this point, I paused the video and asked the LLM, “is the Docker engine a virtual machine?” and a series of questions regarding the engine. I was able to leverage the LLM to a point to understanding the exact nature of the Docker engine, with the answer being, that the engine uses the host OS kernel to hook in resources and execute an abstraction layer.
I then asked it questions about using different operating systems on a custom image and it explained to me how it is done.
These answers cannot be found using Google search. At least, no where near as fast, in depth, and not so specific as an LLM.
These properties of Chat-GPT are invaluable for learning difficult concepts and what is great, is that its trained on nearly all technologies(at least the more popular ones).
Often I ask Chat-GPT questions to help me clarify my notes for my reference manual.
The only downside to Chat-GPT specifically, is, you can’t ask it for the latest advice for the latest releases of frameworks and languages.
Treat the LLM Like Your Personal Tutor
If you are not using an LLM already, you should. You can ask for code examples, for it to update your code and for the model to analyze and explain code line by line.
Afterward, you can ask specific questions about the concepts of the code or framework. I can’t state enough how useful, time saving, and valuable this has become.
This is essentially as valuable as the first time a search engine was created in the 80s & 90s. It was pivotal in how education became accessible to the public. In the same way, in-depth knowledge is now more accessible.
Somehow, magically, I don’t know how, Chat-GPT can teach graphics programming and Machine Learning through questions. How an why it can do this is a mystery to me.
What is better is these topics are notoriously difficult to progress and Chat-GPT significantly lowers the learning curve…
Anywho, I hope you learned something…
Happy coding!
Resources
Example LLM: https://chat.openai.com/
Example Docker chat: https://chat.openai.com/share/a40652a2-eea9-488a-b983-2b5550a3607f
