Tony_Delco

The cost of producing code has dropped to almost zero, but the cost of knowing whether that code is actually useful hasn't.

I think the same is true for images, videos, and music. The cost of producing content is approaching zero, but taste, originality, judgment, and intent still matter.

In a world where anyone can create almost anything, having good taste could become increasingly important.

Minecraft and the Civilization series.

One thing I rarely see discussed is that AI cost is not just dollars per token.

There’s also latency, dependency on external infrastructure, privacy and compliance concerns, energy usage, and just the general predictability of the system itself.

My guess is that this will gradually push a lot of companies toward more hybrid architectures over time. Small or local models are probably good enough for things like filtering, routing or repetitive high volume tasks, while frontier models get reserved for the places where the quality jump actually justifies the added cost and complexity.

As useful as frontier models are, using them for absolutely everything sometimes reminds me of using a distributed system for problems that could have been solved locally with something much simpler.

I wouldn’t be surprised if, in many real world cases, a fast specialized system plus a smaller model ends up being the more practical and economical setup overall.

It’s never too late to learn something new, not even to start in software. Age can bring a different way of looking at problems, shaped by past experiences in different situations and environments.

The path can be tough at the beginning, but if you truly enjoy it, you’re not wasting your time, you’re investing in a new stage of your life.

And this is just my personal opinion: I believe a team should be as diverse as possible. The more different perspectives, the better. A 22-year-old junior doesn’t see things the same way as a 42-year-old junior, and both are valuable and perfectly compatible.

This is amazing work. Most people don’t realize how much effort and depth there is behind writing a TCP/IP stack and plugging it into xv6 with virtio-net and sockets. Huge respect for making it public. Projects like this are gold for those of us who love low-level systems.

To me, the real cost of context switching isn’t measured in minutes lost, but in the break of rhythm. Productivity isn’t about squeezing every second, it’s about getting into a state where your brain runs smoothly, like a warm engine.

When you hop between projects too often, you’re constantly starting cold: reloading mental state, rebuilding assumptions, regaining focus. That’s exhausting, even if each switch only takes a few minutes.

What helped me is not just cutting down on tools, but designing a routine that favors consistency over intensity. I try to dedicate whole blocks, sometimes entire days, to a single project. That way I only pay the switch once, and I get to stay in flow for longer.

In the end, the biggest productivity boost doesn’t come from better apps, but from respecting the brain’s need for continuity.

Congrats on your progress! Looking back, in your earlier projects, what concrete signal would have told you to stop sooner and save time?

With AI, the question isn’t if there’s going to be a bubble… it’s which startups will still be alive when it bursts. Good luck with that.

Review bots are the autopilot. Humans are the pilots. The autopilot will keep getting better, but you still want a human in the cockpit when something truly unexpected happens.

It’s quite a philosophical question. Working eight hours a day for someone else can, in a way, be seen as outsourcing your own will: you think, decide, and act—but always within the framework of what someone else needs. Self-employment, on the other hand, can be deeply rewarding, but it also requires a great deal of self-discipline. In my view, that’s where you run the risk of becoming your own worst boss: demanding, inflexible, and with no clear boundaries.

Fantastic opening line (“Memory oppresses me.”). If this article was written by an AI, it’s the best AI I’ve seen in months.

Seriously though: I already knew the “don’t mix allocators” rule, but I really enjoyed seeing such a careful and hands-on exploration of why it’s dangerous. Thanks for sharing it.

I learned to program in C, and even though I use other languages more often these days, I’m very glad C was my starting point.

It forced me to understand memory, resource management, and what actually happens when you call a function or access a variable. No abstractions to hide behind — just the machine and you.

Even now, when performance matters, I often go back to C-level tools and techniques — mmap, manual memory control, system calls — because they give me the precision I need.

More importantly, learning C gave me fundamentals I don’t think you get as deeply with higher-level languages. It made it easier to pick up other languages, not just syntactically, but conceptually. I understand what they abstract away — and at what cost.

C isn’t perfect, and it’s not the right tool for everything. But if you want to understand computers, I still think it’s one of the best places to start.

And like it or not, there are still things that — for legacy, for portability, or for raw control — can only really be done in C.