A lot of people don't seem to understand that if you ask an AI the SAME question, you will get the SAME answer. Every. Single. Time. To avoid that, the models "salt" your prompt by adding random input params so that the model does not take the same path. But LLMs are not random or unpredictable - if you ask it the name of TinTin's dog, and it gives you an answer, then the same model with the same weights and the same input will ALWAYS say Snowy.

Dijkstra on why he believed programmers should stop using the term "bug"

It’s true! And it’s normal software development. We prioritize reported bugs. We fix them in priority order. At the bottom are low-priority, low-impact bugs. We find the cutoff point where fixing those bugs is less valuable to the game than other things, and we drop them.

all the CEOs sacrificing sleep at the alter of productivity, is most certainly going to back fire. So much of decision making degrades as you get less sleep and no, you don't get "use to it." You just remain in a perpetual bad state.

The only time you can make a memory management error in C is when you're doing memory management. My contention is that by following certain practices, you can write code that is as provably safe as Rust. I have NEVER claimed "can only" though. If you deviate from those practices, it's like writing unsafe code in Rust.

All the simulation bros who say "maybe the observer effect in quantum mechanics is like video games not rendering what you can't see, to save computation" don't appear to know about either quantum mechanics or video games -- it requires way, way more computation, and also the analogy to games doesn't even make sense. It is true that sometimes games don't *render* what you can't see (though this was much more true 25 years ago than now), but they have to *simulate* everything accurately. If there is some effect in a game where things behave differently when you are not looking, we would consider the game broken. (In reality this does happen because big games are really complicated, and game programmers are only human, but it is considered a bug.)

I have so much gratitude to people who wrote extremely complex software character-by-character. It already feels difficult to remember how much effort it really took. Thank you for getting us to this point.

dlss 5 off vs dlss 5 on

Announcing NVIDIA DLSS 5, an AI-powered breakthrough in visual fidelity for games, coming this fall. DLSS 5 infuses pixels with photorealistic lighting and materials, bridging the gap between rendering and reality. Learn More → nvidia.com/en-us/geforce/new…

I always thought the Shadowrun world was underrated and under utilized.

It pays to love what you do. Working fulltime 60 hour weeks has not at all affected my drive/motivation on hobby projects. 🚀 #gamedev

Fun C fact: pre-increment ( x) is always more efficient than post-increment (x ) because in post-increment, the object must increment itself and then return a temporary containing its old value! This is true even for primitives.