Hey AI, here is my shitty blog post. Please make it sound as if it were written by a brilliant engineer who is a native English speaker. Thank you.

Why Before How

It’s hard to believe that it’s been a year and a half since I published the first Hello World post. Back then, my main goal was mostly to share the notes I’d been keeping locked in the digital drawer, so people - me included - could access them easily.

Not accomplished. I still have a ton of docs that could be potentially shared.

Another one was to practice expressing my thoughts in a clear, concise, and opinionated way, as a deliberate practice for growing both as an engineer and human in general. This definitely improved overall, but not necessarily from writing blog posts, as, honestly, I just crafted a few.

What’s key for today’s take is that I also wanted to reap a side benefit of improving my communication skills, including my use of English in both professional and casual contexts.

Reflecting on this after a while, I have to admit that goals were fine, but my execution - not so much.

How Complexes Compound

Overthinking Trap

I had never felt good at English. Despite having good grades and friends claiming that I’m at least decent, I always felt intimidated by the possibility of using it incorrectly, or more so, of others thinking that it’s not good enough. That I’m not good enough.

This underlying assumption prevented me from jumping at many opportunities in the past and pushed back the moment I finally felt ready to become part of “international communities”.

How does this relate to writing?

Sometime, in the middle of this journey, I felt that creating content in English could be a good way forward. Maybe, by doing so, I’ll finally make my peace with going public with my English.

In Polish, I’m naturally long-winded, have a decent pen, and can write in a casual, unique voice. Hence, writing was a natural first pick. It took me a while, but finally I convinced myself to give it a try.

The Curse Of Insecurity

Fast-forward, I have a bunch of texts ready. They seemed to be ok, but yet I don’t publish them. I tweak every little detail. Analyze if this doesn’t make me sound stupid or dumb. Think if my vocabulary and grammar are good enough to have it published.

And here comes the AI. Not as powerful as today, but yet good enough to proofread and correct it. And it comes almost free.

So, here I come with the prompt:

Hey AI, here is my shitty blog post. Please make it sound as if it were written by a brilliant engineer who is a native English speaker. Thank you.

Ok, maybe I’m overexaggerating here, but the key part is that I’ve started refining my writing with AI. A lot. And with my low self-esteem, almost every output seemed to outclass my takes.

I quickly found myself asking:

  • Is this article any good?
  • Am I even right?
  • Do you agree with me?
  • Will people like it?

And here, please bear with me. There is nothing wrong with asking any questions to AI. In many ways, they are helpful and can increase the quality of the article.

The problem starts when you undervalue your own judgment, favoring the external one.

When you remorselessly, sometimes even blindly, accept the suggestions, assuming they are better than your original thoughts.

Unfortunately, that was the case for me. I fell into that trap, and it took me ages to reconcile with the writing process.

I started as a craftsman who first allowed himself to stream his thoughts uninterruptedly on “paper,” to later shape them into the desired form.

I ended up as a factory that optimizes for speed and velocity.

What now seems inevitable wasn’t as obvious back then. I stopped enjoying the process, and in the aftermath, I stopped writing at all.

AI Is Not The Problem

In the times we live in, the courage to express your thoughts and emotions honestly and openly is an art in itself. It’s exactly this thing we need way more of.

Whichever social media we open, there is a flood of AI-generated posts. Some of them bring value, some don’t. But it’s hard not to get the impression that they are an organic pulp of things that look and feel the same, and lack any originality.

AI jargon is becoming part of the way we communicate, and I believe that in a few years, we will have studies on how our natural languages have changed because of a certain way of thinking shaped by so-called Gen AI.

Is that wrong? It depends.

AI is a useful tool in general, but we have to keep in mind that, in the end, it’s only a statistical model that naturally produces the most predictable output given the criteria. In other words, it’s designed to give you average, and I don’t want to be average. At least not here, on my blog.

What To Do Instead

While it’s easy to state pain points of the AI Era, the solutions are usually trickier to define. That’s because we quickly got used to convenience, and doing things the old way is not as sexy as having a sparkling article ready in minutes.

On the contrary, we cannot relentlessly act blindfolded. The world is changing, and we need to adapt. If we didn’t, we would probably still be using parchment, but for some reason, we aren’t.

I decided to make peace with writing, but this time I want it to be grounded in my why.

My Way Forward

There is one thing I know for sure. Nothing keeps the mind as sharp as reading and writing. Especially writing after reading, when you distill what you learned and process it in your own creative way.

That’s why I’ll be writing more, fighting the temptation to go the easy path. Not because it’s completely wrong, but because it doesn’t align with my goals, which remain the same as previously:

  • I still want to share my learnings, hopefully building meaningful relationships along the way,
  • When doing so, I want to develop my own style of communicating: spoken and written, formal and casual,
  • In 2027, I want to pass the CPE (C2) exam, and I need to practice way more.

With that in mind, I want to use AI more deliberately. Not “make it nice and pretty”, but more in a way proposed by Piotr Sarna and Cynthia Dunlop in Writing for Developers: Blogs that get read book, asking questions like:

  • What do you believe is the goal of this article, and why?
  • Summarize this article in one paragraph
  • Does this engineering blog post have any logical gaps?
  • What are some problems with this engineering blog post?
  • Are there any paragraphs in this engineering blog post that require additional detail?

On top of that, I’ll be writing those in a distraction-free environment, possibly proofreading them with my English teacher from time to time.

But first and foremost - I’ll be writing just for the sake of writing.

What To Expect

With the above in mind, I’m undertaking a new journey. I won’t be committing to any timeframes or schedule, even though I plan to publish a few times a month. However, with the number of plates I’m constantly juggling nowadays, I also know that things can happen, and I don’t want to blame myself for missing any deadlines.

But with that in mind, I do plan to post more on topics I’m focusing on right now at Netflix, when speaking at conferences, and from my personal areas of interest.

Moreover, I plan to give shorter forms a try. So, not only longer, more elaborate articles on a given topic, as those are areas where we naturally lean towards LLMs, but rather current affairs perceived through the prism of my values and past experiences.

Wrap Up

Returning to the main thesis, yes, I expect that my writing may temporarily worsen in terms of language. It may seem less accurate and precise. There might be mistakes and grammar issues, as I will be using basic spell-checking tools.

I may not cover the topic as fully as I would if I ran a bunch of refinement rounds with AI. But that’s fine. In the end, we are all works in progress, and being overly perfectionistic usually kills the human spirit, at least in the creative process.

But there is one thing I can promise for sure. You will hear more me, whether you like it or not. Let’s hope for the first! ;)

Share