Tapping the screen for the eighth time, I watch the battery icon at the top right corner tick down from 48% to 28% in what feels like a blink. The device is hot-uncomfortably so-radiating a dry, electronic fever through my palm. I just spent the last eight minutes trying to open a basic banking interface that, for some inexplicable reason, requires a three-dimensional rendering engine just to show me my balance. The frustration is visceral. Earlier, I had to chase a common house spider across the laminate flooring, ending its frantic scramble with a decisive thud of my sneaker. I’m still feeling that jolt of adrenaline, that blunt necessity of force, and it’s coloring the way I look at this unresponsive slab of glass. We are living in a period of profound architectural arrogance, where the digital world is being built exclusively for those who can afford to replace their hardware every 18 months.

I’m thinking about Mia E., a financial literacy educator I met last year who works in a neighborhood where the median income barely touches $28,888. Mia E. doesn’t talk about high-frequency trading or crypto-portfolios; she talks about survival. She told me once about a student who spent 48 minutes trying to load a mandatory government portal on an eight-year-old Android phone. The student was sitting in a public library, tethered to a weak Wi-Fi signal, watching the loading spinner rotate with a mocking, rhythmic consistency. By the time the page finally rendered, the session had timed out. This isn’t just a technical glitch; it is a structural barrier. When we talk about digital reliability, we are often talking about a luxury good masquerading as a utility. The developers sitting in climate-controlled offices in San Francisco or London are testing their builds on M2 Ultra chips and gigabit fiber optics. They don’t see the lag. They don’t see the 18% battery drain caused by an unoptimized JavaScript bundle that serves no purpose other than tracking user ‘engagement’ metrics that nobody actually reads.

It’s a contradiction I see every day. We claim to want a more inclusive internet, yet every major framework update seems to add another 800 kilobytes of weight to the average webpage. We are building Ferraris for a world where most people are still trying to maintain a bicycle. I’ve been guilty of this myself, I suppose. I remember a project back in 2018 where I insisted on using a heavy animation library because it made the transitions look ‘premium.’ I ignored the feedback from the QA lead who pointed out that the frame rate dropped to 8 frames per second on mid-range devices. I prioritized my own aesthetic satisfaction over the actual utility for the end-user. It was a mistake born of privilege, the kind of mistake that feels small in a boardroom but becomes an insurmountable wall for someone trying to pay a utility bill on a lunch break.

This brings us to the core of the frustration: the assumption of the ‘ideal user.’ In the minds of many designers, the user always has a full battery, a recent flagship device, and a stable 5G connection. But that user is a myth. The real user is Mia E.’s student. The real user is the person in a rural area with 2 bars of signal and 8% battery left, desperately trying to find the hours of operation for a local clinic. When code is bloated, it isn’t just slow; it’s expensive. It costs data, which is sold in increasingly predatory packages. It costs time, which is the only resource the working poor have in shorter supply than money. And it costs hardware longevity. If an app is so poorly optimized that it causes a processor to thermal throttle, it is actively shortening the lifespan of that device. We are effectively forcing the people who can least afford it to upgrade their hardware because we are too lazy to write efficient code.

I often find myself wondering if we’ve lost the art of the ‘slim’ build. There was a time when fitting a program into 48 kilobytes was a badge of honor. Now, we pull in a library of 188 dependencies just to center a button. The inefficiency is a form of environmental and social decay. It’s the digital equivalent of a gas-guzzling SUV idling in a school zone. We need to start demanding better. We need to look at platforms like ems89 and ask why more of the web can’t prioritize that level of lean, functional performance. Reliability shouldn’t be a premium feature that you unlock with a $1,388 purchase at a glass-fronted retail store. It should be the baseline. It should be the promise.

There is a certain violence in bad software. It’s a quiet, bloodless violence, but it’s there. It’s in the sigh of a mother who realizes she can’t access her child’s school portal. It’s in the frantic tapping of a screen that refuses to acknowledge a touch command. When I killed that spider earlier, it was a sudden, messy end to a small life. Bad software is a slower death-a death of a thousand cuts, or rather, a death of a thousand dropped packets. It erodes trust. It makes the digital world feel like a place where you are only welcome if you have the right credentials and the right silicon in your pocket.

Mia E. once showed me a collection of ‘dead’ phones in her classroom-devices that were perfectly functional in terms of hardware but had been rendered useless by software updates that they couldn’t handle. There were 28 of them on a shelf. They looked like artifacts from a lost civilization. ‘These aren’t broken,’ she said, ‘they’ve just been outpaced by vanity.’ That phrase stuck with me. Vanity. The vanity of the developer who needs the latest shadow effect. The vanity of the product manager who needs every single user action to be logged in a real-time database. We are drowning in vanity, and the weight of it is pulling down everyone who doesn’t have a life jacket.

I’m not saying we should return to the text-only web of 1998, though some days that sounds like a paradise. I’m saying that we need to acknowledge the reality of the hardware gap. If your application doesn’t run smoothly on a device from 8 years ago, you haven’t built a tool; you’ve built a gate. And that gate is locked for a significant portion of the population. We need to start treating CPU cycles as a finite resource, something to be spent with care and respect for the user’s battery life. Every time I see a site take 8 seconds to load, I think about the energy being wasted globally. I think about the millions of devices heating up, their batteries degrading just a little bit faster, all for the sake of some bloated tracking script or a high-res video background that nobody asked for.

It takes a certain amount of humility to strip things back. It’s much easier to add another layer of abstraction than it is to dig into the core of a performance bottleneck. But that work is necessary. It’s the difference between a tool that empowers and a tool that excludes. As I sit here, my phone finally cooling down after that 8-minute struggle, I feel a sense of exhaustion that has nothing to do with the physical world. It’s the mental fatigue of fighting with tools that are supposed to serve me. I look at the spot on the floor where the spider was, now clean but for a faint mark, and I realize that we spend so much of our lives squashing bugs-both literal and digital-that we’ve forgotten how to build environments where they don’t thrive in the first place.

If we want a digital future that actually works for everyone, we have to stop building for the 8% of the population with the newest tech. We have to start building for the 88% who are just trying to get through their day on a device that’s seen better years. We need to stop seeing ‘performance’ as a technical metric and start seeing it as a human right. Because when the screen lags, and the battery dies, and the portal fails to load, it’s not just the code that’s broken. It’s the connection between the person and the progress we promised them. Mia E. shouldn’t have to explain to her students why they can’t access their own financial data. I shouldn’t have to watch my battery drain 18% just to check a balance. We can do better, but only if we decide that reliability is more important than the latest shiny, heavy, useless trend. The question is whether the people holding the keyboards actually care about the people holding the phones, or if they’re too busy staring at their own high-resolution reflections to notice the lag.