Working 16 Hours a Day and Still Feeling Behind in AI

I have been working 16 hours a day, 7 days a week. And I still feel behind.

Not behind on a deadline. Not behind on a client. Behind in something harder to name: the sense that the world is moving faster than I am, even when I am moving faster than I ever have. It is a strange place to be. Productive and anxious at the same time. Building real things and still feeling like you are missing something you cannot quite identify.

I am writing this because I think a lot of people in tech are sitting with the same feeling and not saying it out loud. This is my attempt to say it out loud.

What the Days Actually Look Like

I wake up before the alarm. Brain already running. By 7am I am at the screen.

Mornings go to WP Vanguard. It is a WordPress security scanner I have been building because I kept watching site owners get wrecked by malware they never saw coming. They would discover the problem weeks after it happened, after Google had already flagged the site, after the damage was done. I knew I could build something that caught it earlier and I did. Scanner, cleanup workflows, backend infrastructure, frontend, onboarding, pricing, the whole product layer, mostly just me and Claude moving fast through decisions that used to require a team.

There is a specific kind of momentum that happens when you are building something you genuinely believe in and the tools are finally fast enough to match your thinking. That is what mornings feel like right now. I sketch a feature, Claude helps implement it, I test it, I push it. The feedback loop is tight. Things that would have taken me a week two years ago take a day now. That part is genuinely exciting.

What is harder to describe is the decision overhead. Every day there are ten things I could work on that would all be legitimate choices. Which scanner feature matters most to the next hundred users? Should the cleanup workflow handle this edge case automatically or flag it for manual review? Is this the right moment to build the API layer or should that wait until the core is more stable? These are not questions that automation solves. They require judgment, and judgment requires actually being in the problem deeply enough to have a feel for it. That is where the hours go. Not just doing but deciding.

Afternoons shift to Wbcom Designs. Over a hundred plugins and themes, tens of thousands of users, and someone always needs something. A bug surfaces. A WordPress update breaks a plugin that was perfectly stable last week. A customer has been waiting two days for a response and the queue is backed up. I have a team and they are good, but I stay in the weeds. I always have. I think some part of me is afraid of what happens if I stop knowing the products deeply enough to debug them myself.

In between all of that I am also building the infrastructure layer. I spent months this year building out automations: marketing, blog scheduling, support triage, QA workflows. Each one solved a real bottleneck. Each one was the right thing to build at the time. The support triage system alone saves hours every week. But here is the thing nobody tells you: when you automate a problem, you do not get that time back. You spend it on the next problem that was previously below the line of what you could get to. The queue of things I could be doing expands to fill whatever capacity I create. The list does not shrink. It regenerates.

Evenings I read. AI papers, release notes, discussions on X and Hacker News about where things are going. Not because someone told me to. Because I find it genuinely interesting and because some part of me is always trying to close the gap between what I know and what I feel like I should know. By midnight I am usually still at the screen. Sometimes on something urgent, sometimes just on the thing I could not stop thinking about at 10pm. Then I wake up and do it again.

Every system I build to free up time just fills with the next thing. The list never shrinks.

The Part I Usually Skip

My family thinks something is wrong with me.

That is not a throwaway line. From the outside, this is what they see: someone physically present but mentally elsewhere. At the dinner table but solving a problem in his head. Hears you, responds, but is clearly half-listening to some internal loop about a feature decision or a support backlog or a thing he read earlier that he cannot stop turning over. My wife has mostly stopped asking what I am thinking about. She knows the answer.

I do not want to be that person. I know what it costs. I have watched other builders become their work in a way that hollows out everything adjacent to it. I tell myself I am not doing that. But the honest version is: I do not fully know. It is hard to see clearly from inside it.

The 16-hour days sound impressive or alarming depending on who you are telling them to. But neither reaction captures what they actually feel like from the inside or what they actually cost. The hours are not the point. The point is that work has become the default state and everything else has become the interruption. I am physically in the room and mentally somewhere else, and the people around me feel that even when I do not say anything.

I built automation that handles blog posts, support tickets, and QA checks. None of it has automated presence. That problem does not have a system design solution.

The Contradiction I Live In

Here is what makes it complicated: I genuinely love what I am building.

I am not grinding through misery. WP Vanguard is a product I am proud of. When a site owner runs a scan and it catches something serious before it does real damage, that matters. That is real value in the world. The plugins at Wbcom are used by real communities. The automations I built this year made my team’s work meaningfully better. When I look at what I actually shipped, it holds up.

And yet, every single week, something ships that changes the landscape. A new model with a capability that did not exist seven days ago. A company of ten people building in a month what I thought was a year of work. A tool that makes part of what I built last quarter feel like a workaround. The pace is not slowing down. If anything it is accelerating. I am building fast and the feeling of being ahead never arrives.

That is the treadmill. Not that I hate running. That the treadmill keeps getting faster and the feeling of arriving somewhere never comes. You are moving, you are working, you are shipping, and somehow the gap between where you are and where the field is does not close. It just stays there, humming in the background.

The treadmill feeling also does something specific to how you make decisions. When you feel behind, you start optimizing for coverage over depth. You want to know a little about everything that is happening rather than a lot about the things that actually matter to what you are building. You read more and absorb less. You try more tools without committing long enough to really learn any of them. The anxiety produces a kind of scattershot attention that is actually the opposite of what the situation requires, which is focused, deep work on the things most likely to matter. I notice this in myself and I correct for it imperfectly.

What I keep coming back to is that these two things, loving what you build and feeling behind in the broader race, are not mutually exclusive. I thought they were, for a while. I thought if I was proud of the work then the anxiety would resolve itself. It does not work that way. You can be doing real work and still feel the ground shifting under you. You can be moving fast and still feel slow relative to something larger than you.

I love what I am building. And I still feel like I am losing a race I did not choose to enter.

The Fear Underneath

There is a voice that runs underneath all of this. Ye bhi kar le, wo bhi kar le. This too. That too. It is not ambition exactly. Ambition has a direction. This is more like fear with no clear target. The fear that the window to evolve is narrowing and I am not moving through it fast enough. The fear that by the time I figure out where to go, the opportunity to go there will have closed.

And the list of things I am moving toward keeps growing. WP Vanguard is not the only thing in the pipeline. There is Mediaverse, a media management product, coming in free and pro versions. WP Sell Services, for service-based businesses on WordPress. WP ConnectPress. Career Board, also in free and pro. Each one of these is a real product solving a real problem. Each one has been thought through. But they all exist simultaneously in various stages of being built, being planned, being positioned. That is not a complaint. I chose this. But the breadth of it is also a direct answer to the question of why the days are 16 hours long and still feel short.

Every major technology shift has a window. There is a period where the people who learn the new paradigm early and deeply get positioned in ways that last. They build intuition when it still matters to have intuition. They build products before the space gets crowded. They build reputations before everyone else has one. And then the window narrows, not because the technology stops evolving, but because the early advantage fades as everyone else catches up.

I watched it happen with WordPress. There was a window in the early days where people who understood it deeply built businesses and reputations that carried them for a decade. I got in. Not at the very beginning but early enough. I learned the platform at a level that mattered. The window was open and I went through it. That feels like luck in retrospect, or timing, or just the fact that I was paying attention at the right moment.

With AI, I genuinely do not know where the window is. I use these tools every day. They are embedded in how I build WP Vanguard, how I run support, how I write, how I think through problems. But using something well and understanding it at an architectural level are not the same thing. I am a product builder, not an ML researcher. I operate at the application layer, not the model layer. And some part of me keeps wondering if that means I am permanently one abstraction away from where the real understanding lives.

I wrote earlier this year about how AI won’t kill web development agencies, it will kill the bad ones. I still believe that. The threat is not the technology itself. The threat is standing still while the technology moves. But believing a frame does not make the daily uncertainty quieter. At midnight, when I am reading about some new capability and trying to figure out what it means for what I am building, the frame does not do much work.

What actually sits underneath the 16-hour days, I think, is this: I am trying to be everywhere in the hope that one of the places I am will turn out to be the right one. The scanner, the plugins, the automations, the content, the reading. It is a broad surface area and I cannot fully defend all of it. But narrowing down feels like a risk I cannot take when I do not yet know which bets are the right ones.

I also think there is something specific happening with the pace of AI releases that makes this worse in a way that did not apply to earlier technology cycles. When WordPress was evolving, the pace was slow enough that you could keep up by reading the release notes and testing things. A major release happened once or twice a year. You had time to absorb it. The pace of AI development is different in kind, not just in degree. There is no week where nothing significant happened. There is no month where you can say “I am fully current.” The target moves every few days. That is genuinely new. The anxiety it produces is not irrational. It is the correct response to actually keeping track of how fast things are moving.

One Thing That Has Helped Slightly

I have started trying to separate two races that I keep running together.

One is someone else’s race: understanding AI at the model level, being at the frontier of capability, publishing research, building foundation models, being the person in the room who predicted where things would be eighteen months from now. That race exists. Real people are in it. It is a legitimate and important race. I am not in it. I was never going to be in it. And measuring myself against it was making everything worse.

The other race is mine: understanding AI well enough to build better products for the people who actually use what I build. Using what exists to do things that would have required a much bigger team two years ago. Learning the shape of what is possible so that my products improve over time. Staying useful and relevant to the actual users in front of me, not to some abstract definition of the frontier.

WP Vanguard was not built by someone who understands transformer architectures. It was built by someone who understands WordPress deeply, understands security, and understands what a site owner is actually afraid of at 11pm when they get an alert. I used AI to build faster and better than I could alone. The product exists. It helps people. That was the right race to run, and I ran it.

Separating these two races does not make the anxiety disappear. The voice that says “ye bhi kar le, wo bhi kar le” does not go quiet just because you have a cleaner frame for what you are doing. But it makes the work feel less like running away from something and more like running toward something. There is a version of this I keep coming back to in product decisions too: build what the market needs, not what you feel like you should be building. The same instinct applies to learning and positioning. Stop trying to run a race you were never in, and run yours with more intention.

There are two races. I keep confusing them. Only one of them is mine to run.

What I Have Not Figured Out

I want to be careful not to make this sound more resolved than it is.

I do not know if the window is still open. I do not know if product-layer builders are well-positioned for what comes next or if the ground is going to shift in a way that changes the value of everything I currently know. The history of technology is full of people who were experts in the thing that got replaced and could not fully make the transition to the next thing. I do not know if I am in that category yet. I hope not. But I cannot rule it out, and that uncertainty is real.

I do not know if 16 hours is the right number or if I have just convinced myself it is necessary because stopping feels more dangerous than continuing. There is a version of the grinding that is productive and a version that is just fear in disguise. I am not always sure which one I am doing on any given day.

I also do not know what the alternative looks like in concrete terms. I have tried to imagine doing fewer hours and I cannot see what I would cut. The WP Vanguard work needs to happen if the product is going to exist. The Wbcom work needs to happen because there are users depending on those products. The reading needs to happen because without it I genuinely do not know what I am missing. Every thing on the list feels load-bearing. Maybe some of it is not. But I have not yet been able to identify which parts. So instead of cutting, I just keep fitting more in, which is how you end up at 16 hours without ever making a single decision to work that much.

I definitely have not figured out how to be present with my family. The support tickets triage themselves now. The blog posts schedule themselves. The family dinner still happens at a table where part of me is somewhere else, and no amount of automation changes that. That is a real problem and I do not have a good answer for it.

I also have not figured out what “keeping up” would even look like if I achieved it. What is the state where I wake up and feel like I am not behind? I genuinely cannot picture it. That suggests the feeling might not be about the gap at all. It might be structural. Something in how I relate to the work, or to progress, or to uncertainty itself. That is an uncomfortable possibility because it means the solution is not more hours or better tools. It is something harder to name and harder to change.

What I do know is that this feeling is not mine alone. I see it in how other builders talk when they are not performing confidence for an audience. The specific texture of working hard, loving the work, and still feeling like you are losing ground. That is not a personal failure or a sign of doing it wrong. It is an honest response to a genuinely strange and fast-moving moment. The pace is real. The uncertainty is real. Acknowledging both seems more useful than pretending one of them is not there.

---

I am writing this from the middle of it, not from the other side. Still at 16 hours. Still loving the work. Still uncertain about the gap. Still figuring out the family part.

If you are in it too, I would like to know what it looks like for you.