Reaching Top Ranks on HTB: What Changed Along the Way

After 6 or 7 years on Hack The Box, competing, failing, learning, and coming back again, I finally reached the Top 7 worldwide. The number itself is not what matters most, but everything behind it: time under pressure, repeated failure, small improvements, and a long sequence of problems that forced me to think deeper than I initially wanted to. This is not a victory lap, it is more of a checkpoint, one of those moments where you stop briefly, look back, and try to understand what actually changed along the way.

What Years of CTFs Actually Build

People often see rankings and assume speed or talent, but most of it comes down to endurance and how you deal with being stuck. You spend hours on things that look trivial, chasing wrong assumptions, reading documentation that does not immediately help, and trying approaches that fail quietly without giving you feedback. Over time, something shifts in your mindset, and you start recognizing patterns faster, but more importantly, you start questioning your own assumptions earlier. That ability to stop, reassess, and challenge your own thinking is far more valuable than simply knowing more techniques or memorizing payloads.

Back then, solving challenges required a different kind of patience, because you were forced to read a lot and often without clear direction. Documentation, source code, random blog posts, even RFCs became part of the process, and many times you did not even know what you were looking for, just that something was missing. That process was slow and sometimes frustrating, but it forced your brain to build connections, to explore ideas, and to develop a deeper understanding of systems. It was inefficient in terms of time, but very effective in terms of long term learning.

The Shift: AI Changed the Workflow

There is no way around it, AI changed that workflow completely, and pretending otherwise is just ignoring reality. What used to take hours or even days can now be reduced to minutes if you ask the right questions and guide the model properly. Some people believe this killed the essence of CTFs, while others fully embrace it without questioning the tradeoffs, and I find myself somewhere in the middle. There is clear value in reducing friction and accelerating problem solving, but there is also something important that gets lost when that friction disappears entirely.

The real difference today is not the existence of AI, but how each person decides to use it in practice. You can treat it as an assistant that helps you reason faster, validate ideas, and explore alternative approaches while still understanding the context and the underlying mechanics. Or you can outsource the entire process, copy the challenge, get an answer, verify that it works, and move on without learning much from it. Both approaches can lead to solving a box, both will give you the same visible result, but only one actually builds something inside your head that you can reuse later.

What Gets Lost Without Friction

The dangerous part is that outsourcing still gives you a sense of progress, because you get the flag, you complete the challenge, and you feel like you are improving. In reality, you might not even understand five percent of what just happened, and that gap becomes a problem when you face something slightly different. Without understanding the assumptions, the constraints, and the reasoning behind a solution, you are left with fragments instead of knowledge, and that is where many people get stuck without realizing it.

Before AI, you were forced to sit with problems longer than you wanted to, and that discomfort was part of the training. You would misinterpret outputs, try incorrect payloads, spend hours debugging something simple, and eventually step away only to come back later and see what you missed. That cycle trained your ability to deal with uncertainty, to persist without immediate reward, and to explore without guarantees, which are all skills that do not show up in rankings but define how you operate under pressure.

A Note for the New Generation

If you are starting now, you have access to tools that dramatically accelerate your learning curve, and that is something worth taking advantage of. At the same time, it becomes your responsibility to not lose the parts that actually matter, like attention, curiosity, and the ability to understand systems beyond surface level results. When AI gives you an answer, it should not be the end of the process, it should be the beginning of a deeper analysis where you ask why it works, what assumptions it relies on, and how it could fail.

Learning from AI requires intention, otherwise it just becomes a shortcut that limits your growth. Treat it as a guide, not as a replacement, and force yourself to understand the context behind every solution you accept. That discipline is what will separate people who actually improve from those who only move faster.

Closing

Reaching Top 7 is a milestone, but more than anything it is a reminder that tools will keep changing and workflows will continue to evolve. What remains valuable is your ability to think clearly, question assumptions, and build a mental model of how things actually work. Speed is useful, but understanding is what stays with you when things break, when outputs are incomplete, and when there is no obvious answer.