Why interfaces that act human make us misread what machines actually do — and why that misreading is now a design choice with real costs.
SULAIMAN ALSHAMMARI
NAMED & REACHABLE · REALITY ONLY
Network technician, CCNA-certified, with three Linux Foundation certifications. I prefer a local-first, hybrid workflow — owning my tools outright rather than renting them indefinitely through subscriptions.
I build with C, C++, C#, .NET, Java, and Assembly.
Most tools marketed today as "AI" are advanced statistical prediction layers, not genuine intelligence. Conversational interfaces are designed to feel human — first-person pronouns, apologies, confidence — but that design choice encourages users to attribute understanding, intent, and reasoning to a system that is, underneath, recombining patterns from its training data.
That gap between how it feels and what it actually does isn't an accident. It's a UX decision, and it has consequences: users trust outputs they shouldn't, defer judgment they shouldn't defer, and mistake fluency for correctness.
My interest is in closing that gap — not by making machines seem more human, but by being honest about what they are, and building toward systems capable of genuine reasoning rather than persuasive mimicry.
C · C++ · C# · .NET · Java · Assembly · Linux
Riyadh, Saudi Arabia 🇸🇦