first, to explain why i prefer writing over other forms of communication, you must understand how my brain works. and to do that, i need to give you a glimpse of what my mind looks like 99% of the time:
beautiful right?
if you still don’t understand, let me explain: at any given moment, my mind is a jumble of tangled thoughts, making it difficult to follow a single line of reasoning. it’s hard to process what i’m thinking, figure out what i need to say next, or even decide whether i like something. if the main idea i need to focus on is subjective, there’s a good chance my brain will freeze. believe me, it’s chaotic. but don’t get me wrong: i’m fully functional. that’s why i prefer writing, because i don’t really mind how much time it takes.
the second reason is that i need to improve my english writing skills. after i finish writing, i look for corrections and ways to improve, and by doing this, i get better. make sense?
the third reason is to reinforce my learning by trying to share what i’ve learned with others.
finally, the last reason is that i’ve come across so many poorly written posts on the internet, especially on linkedin. many are generated by chatgpt or other language models, while others spread misinformation about nlp, or worse, they misrepresent basic fundamentals of machine learning.
i’ll be honest: i don’t aspire to become an ai influencer, but i do want to be a recognized reference in ml/dl/ai/nlp (lol). and i have no intention of achieving that by pretending to be someone i’m not, or haven’t become yet. the most cringe-worthy thing i’ve seen lately in the ai field is professionals giving themselves titles or claiming expertise they haven’t earned due to a lack of experience.
that said, i want to contribute to the community, especially because i see a significant gap that needs to be filled, particularly for data scientists when it comes to software engineering knowledge. i’ve worked with many peers, and it’s surprising how little interest they show in this area. they have a deep understanding of data science, but their implementations and models often create serious bottlenecks and the cost of that can be very high. a well-implemented basic model is far more valuable than a high-accuracy model that performs like a clunky cart during inference.
ultimately, the goal is to provide clear explanations, tutorials, demystified theories, and studies. so, i would greatly appreciate any feedback, corrections, or encouragement you can offer.
see you soon!
ps.: is cluck cart a good translation for “carroça”?