blog

day 5 was supposed to be simple: count ids in ranges. beating a 0.4ms target meant ditching clever data structures for primitive arrays, cache-friendly loops, and a JVM warmup trick.

day 3 of advent of code 2025 was less about algorithmic breakthroughs and more about squeezing every last millisecond out of python. levi had a 1.4ms solution on part 1. i was determined to beat it.

day 2 of advent of code 2025 had me learning lua from scratch while discovering that sometimes the best optimization isn't faster code. sometimes it's just realizing you're solving the wrong problem entirely.

day 1 kicked off this year's advent of code showdown in php with a deceptively simple puzzle that taught me a painful lesson about testing, debugging, and how bugs can cancel each other out.

my computer science program in the netherlands is a major regret. it's superficial, often requiring ai for projects instead of teaching me to code. though i'm financially trapped, i'm pushing through by self-learning and building things, because creating is the only genuine and motivating part of my programming journey.

a thought on the current ai hype, especially in programming. while tools are making coding more accessible, i'm looking at the potential downsides of "vibe coding" and how relying too much on ai might affect the critical thinking and unique problem-solving i see as the art of programming.

reflecting on my ongoing computer science studies, i realized that while the curriculum covered fundamentals, it often lacked practical application, and many peers relied on ai tools instead of truly learning. i now prioritize building real projects and independent learning, finding it a more rewarding way to develop problem-solving skills and advising new students to do the same.