You don't need to write tutorials*

Ivan G

Sep 20, 2024 — 2 min read

The majority of them could be obsolete.

And here is why.

*️⃣

If it's trivial and contains information that is already present in thousands of other tutorials, perhaps it would be better to write a tutorial featuring a new, unique technique or research.

If nothing comes to mind right now, consider sharing your thoughts or experience. 👾

The paradigm has just shifted, and most of us didn't notice it yet.

We all have to acknowledge that the normal flow to search for precious data bits — is going to be deprecated.

What now? We are steadily switching to LLMs as our search engine for information.

And of course, currently they do have the problem with reasoning. It will lie to you right in the face, and it won't even blink. Perplexity seems to partially fix the issue, and at least it will give you the sources that you could check manually. And if you care about the topic you are asking, then you should definitely reinspect the sources.

But you know: you don't need classic, fundamentals tutorials that are being repeated and rephrased by a lot of individuals and organizations around the world.

Need to learn Kafka? LLMit.

Struggling to deep dive into AWS infrastructure? LLMit.

Trying to translate one SQL dialect into another? LLMit. Or actually, use sqlglot :)

💡

You don't need the basic tutorials much if you can just ask for a cheatsheet or quick-start guide from LLM. They've already parsed the majority of publicly available tutorials anyway ¯_(ツ)_/¯

And the killer feature for me is that LLM could actually adapt the tutorial for your environment. Isn't it crazy? It's like having a highly specialized PhD who may adjust for your needs!

I know it may sound like an utter LLM propaganda. But here comes the asterisk: if you actually got the novel idea, creating your own stuff, doing the research, and the content of your tutorial isn't in major LLMs, then it would be a great contribution to the humanity knowledge library!

But please, think twice if you want to write another Python "Hello World" tutorial. If there is nothing to highlight which will make your tutorial stand out aside from 1000+ of the similar tutorials, maybe you could find something unique?

Humanity really needs unique things. Or at least quality ones. Ordinary things have a great chance to be drowned in the river of information.

Stay geniune. Make geniune.

DYOR as always.