i'm writing for the person i was at 2am

The average attention span on a single screen is 47 seconds.

I need you to really sit with that for a second. Forty-seven seconds. In 2004 it was 2.5 minutes. That's an 80% decline in twenty years. Gloria Mark at UC Irvine has been tracking this and the line only goes one direction.

The median is even worse: 40 seconds.

Meanwhile, I'm out here writing 3,000-word blog posts and expecting people to read them.

This is either incredibly stupid or the smartest thing I've ever done. I genuinely cannot tell. But I have a theory, and I think the data backs it up.

Let me give you the brain rot numbers first because I think the scale is important.

The average TikTok user spends 95 minutes per day on the app. That's 1 hour and 37 minutes. They open it 19 times per day, with sessions averaging about 11 minutes each. 48.5 hours per month. On one app.

21% of American teenagers report using TikTok "almost constantly." that's not a survey artifact. that's a fifth of an entire generation with a dopamine IV drip running through their pocket.

Researchers at King's College London found that heavy short-form media use is associated with abnormal white matter in the brain - specifically in regions linked to behavioral control - and a 15% shorter sustained attention span. This isn't "kids these days" complaining. This is neuroimaging.

US teens average 8.5 hours of screen time daily. Eight and a half hours. That's more than a full workday. That's more time than they spend sleeping.

And the content they're consuming during those 8.5 hours is optimized, ruthlessly, algorithmically, with billions of dollars of engineering behind it, for one thing: keeping you scrolling. Not making you smarter. Not giving you something to think about tomorrow. Keeping your thumb moving for one more second. And then one more. And then one more.

The word for this - and I'm not being dramatic, Oxford literally chose it as their word of the year - is brain rot.

So. Into this landscape. I am publishing multi-thousand-word essays about context transfer and productivity psychology and my immigrant parents not knowing what a startup was.

Why?

Here's the thing nobody in the content world wants to say out loud because it would invalidate most of their business models:

Short-form and long-form content are not competing for the same audience.

They're not even competing for the same type of attention. They are two completely different activities that happen to exist on the same internet.

Short-form content captures reflexive attention. The scroll. The swipe. The 3-second hook that fires up your dopamine system before your prefrontal cortex even registers what happened. It's a slot machine. You pull the lever, you get a random reward, you pull again. TikTok knows this. Instagram knows this. Every platform optimized for engagement over time-spent knows this.

Long-form content captures intentional attention. You don't accidentally read a 3,000-word essay. You choose to. You're in a specific state of mind - curious, searching, hungry for something that the scroll isn't giving you. You're not killing time. You're investing it.

These are fundamentally different psychological states. And they attract fundamentally different people.

Or more precisely - they attract the same people in fundamentally different moments.

Here's where the economics get wild.

YouTube Shorts generate roughly $0.18-$0.20 RPM (revenue per thousand views). Long-form YouTube content generates $3.33-$5.50 RPM. that's a 15-30x difference in the value of each viewer's attention.

Read that again. The person watching a 12-minute video is worth fifteen to thirty times more than the person watching a 30-second short. Not 2x. Not 5x. Fifteen to thirty x.

Why? Because the person watching the long video is demonstrating intent. They're choosing to be there. They're the kind of person who makes purchasing decisions, who has budget authority, who reads the whole email before responding. The person swiping through shorts might be doing it on the toilet. (you know I'm right.)

It's the same story with newsletters. The average newsletter subscriber generates $5.23-$7.26 per year in revenue. The average YouTube subscriber generates $1.52-$2.45. Substack just hit 5 million paid subscriptions - up from 2 million in 2023. The top 10 Substack creators collectively earn $40 million per year. 52 newsletters earn at least $500k annually.

The people paying $10/month for a newsletter are not the same audience as the people doom-scrolling reels at midnight. Or rather - they might be the same humans - but they're in an entirely different mode. One mode generates $0.20 per thousand. The other generates $5-7 per subscriber. The arbitrage is enormous and almost nobody is exploiting it because everyone is chasing the big number (views) instead of the valuable number (intent).

Now let me tell you who's actually reading long-form content in 2026.

75% of executives cite email newsletters as their primary information source. Not Twitter. Not LinkedIn. Not TikTok. Newsletters. Curated, long-form, intent-driven content delivered to their inbox.

C-suite attention peaks before 8:00 AM. They're reading while their coffee brews, before the meetings start. Once 9AM hits and the calendar takes over, content consumption drops by more than 80%. That window - 6 to 8:30 AM - is when the most powerful people in your industry are actively looking for something worth reading.

And here's the kicker: 70% of B2B buying decisions are made before someone ever contacts sales. The research happens in the dark. Executives consume 3-7 pieces of long-form content during this silent evaluation phase. White papers. Deep blog posts. Detailed case studies. Not carousels. Not 30-second clips. The things that demonstrate you actually understand the problem at a level of depth that can't be faked in a tweet.

Executive-led content on LinkedIn earns 8x more engagement than brand-page content. Because the founders and leaders who actually write - who actually think in public - signal something that a marketing team running a content calendar never can: this person has taste, and this person has depth, and this person has actually done the thing they're talking about.

I have a confession that connects all of this.

I have a Readwise account with over 10,000 saves.

Ten thousand articles, essays, blog posts, research papers, Twitter threads, and random corners of the internet that I highlighted, saved, and tagged over the last several years. I might release a search over it someday. For now it's been my second brain.

But here's the thing - I didn't curate 10,000 saves during the day. I curated them at 2am. At 3am. In the state of mind that nobody optimizes content for: the deep exploration mode.

When I was 17, scrolling Reddit and Stack Overflow trying to figure out what computer science even was because nobody in my immigrant family could tell me - I wasn't looking for a carousel. I wasn't looking for "5 tips to break into tech." I was looking for signal. Real signal. The kind of thing that makes you stop scrolling and actually read.

I was looking for the post that felt like someone was talking directly to me. That understood the specific confusion I was sitting in. That didn't try to be everything to everyone but instead went so deep on one thing that if it was the right thing for you, it felt like finding water in a desert.

I found maybe a hundred of those posts across thousands of hours of searching. And each one changed something in how I thought. Shaped a decision. Reframed a problem. Made me feel less alone in the chaos of figuring it out without a map.

I wrote about this recently - about being the kid that nobody showed how. About Reddit at 2am and cold emails and building my own roadmap out of scraps. That piece was 3,000 words. It violated every content rule. No hook in the first line. No listicle. No "here are my 7 takeaways." Just a long, winding, honest walk through what it felt like to figure it out alone.

I wrote it for the person I was at 2am.

Here's my theory and I think it's right:

The content landscape is bifurcating. Hard. And the middle is dying.

On one end: algorithmic short-form content optimized for maximum reflexive engagement. Views in the millions. Revenue per user approaching zero. Audience: everyone, in their lowest-intent state.

On the other end: deep, long-form, intent-driven content that will never go viral. Audience: small, incredibly high-value, in their highest-intent state. The founders pre-launch. The executives before 8am. The college kid at 2am who is going to build something important and is looking for the signal that tells them they're not crazy.

The middle - the "pretty good" 800-word blog post, the informative-but-forgettable LinkedIn post, the serviceable newsletter that recaps the week's news - that's dying. Not because it's bad. Because it's not deep enough to capture intentional attention and not short enough to capture reflexive attention. It sits in no-man's-land.

I'm not fishing in the big pond. I know this. My essays will never get a million views. They won't trend on Twitter. They won't get picked up by the algorithm and shown to people who didn't ask for them.

I'm fishing in the small pond. The 2am pond. The "this person is actively searching for something and will read every word if I say something real" pond.

And here's why I think that's the better pond:

Because the people in it are the ones who actually do things.

I keep coming back to something.

68% of readers "skim first" to decide if a piece is worth a deep dive. They'll give you 3 seconds. If you fail the scan, they're gone. That sounds like an argument for shorter content.

But I think it's actually an argument for deeper content. The 68% who skim and leave? They were never going to read it anyway. They were in reflexive mode. The 32% who scan, decide it's worth it, and actually read? Those are the people who matter. And they can tell - instantly - whether something was written for them or written for the algorithm.

The algorithm wants you to optimize for the 68%. Hook in the first line. Carousel format. Bite-sized takeaways. "Save this for later" bait.

I'm optimizing for the 32%. And within that 32%, I'm really optimizing for the 5% who will read every word, think about it for a week, and send it to one friend with a note that says "you need to read this."

That's how ideas spread in the high-value pond. Not through algorithmic distribution. Through one person sending it to one other person who they thought needed to see it.

Paul Graham has been doing this for 20 years. His essays are 5,000 words. No images. No SEO optimization. No social media strategy. Just deep thinking published on a website that looks like it was designed in 1997. And they've influenced more startup founders than any content marketing operation in history.

That's not a scalable strategy in the traditional content sense. But it's the most valuable content strategy that exists. Because when one Y Combinator founder reads a PG essay and it changes how they build their company, that's worth more than 10 million TikTok views.

The value of a reader is not equal across readers. And the readers who consume long-form content are, on average, dramatically more valuable - as customers, as collaborators, as signal-boosters - than the ones who consume short-form.

I wrote "context transfer is dead" last week. It's about 3,000 words. It starts with a spreadsheet version history and ends with a thesis about process literacy being the new great divide. If you're in reflexive mode, you bounced after the first paragraph. If you're in the 2am mode - the searching mode, the "I'm trying to understand what's actually happening with AI" mode - you read the whole thing.

I wrote "three sentences that governed my entire life" the same week. It's longer. It walks through three quotes that defined three phases of my life. No listicle. No "here's what you can learn from my journey." Just the raw pattern I noticed while writing previous pieces, caught in real time, with the full emotional texture intact.

I wrote both of those in under 30 minutes. Not because I'm a fast writer. Because I built a workflow where I do the discussion and planning and the AI does the drafting and research. I talk to it about what I'm thinking. It probes me with hard questions. By the time I've answered them, I understand what I actually want to say. Then it drafts for me in my voice because it's been learning my voice for months. And I review.

This is the irony that I think about constantly: I'm using AI - the thing that's arguably accelerating brain rot by making content infinitely producible - to write the kind of content that is the opposite of brain rot. Deep, long, intentional, honest content that demands 15 minutes of someone's undivided focus.

The tool doesn't dictate the output. The taste does.

And taste is the thing that can't be automated. It's the thing that separates the person who uses AI to generate 47-second content from the person who uses AI to write the essay that changes how someone thinks about their career at 2am.

So here's what I'm doing, explicitly, and why.

I'm writing long. I'm writing deep. I'm writing in a way that violates every "best practice" that every content strategist will tell you. No hooks. No formatting tricks. No "TL;DR at the top." No concessions to the 47-second attention span.

Because I'm not writing for the 47-second attention span. I'm writing for the person I was. The 17-year-old on Reddit at 2am who was desperate for signal and drowning in noise. The first-gen kid who couldn't name the career he was reaching for. The builder who knew they were supposed to build something but couldn't find anyone who'd written honestly about what that actually felt like.

That person will find this. Maybe not today. Maybe not through the algorithm. But through a friend who sent a link. Or a Google search at 3am. Or a Readwise save that resurfaces six months from now at exactly the right moment.

That's how the good stuff always found me. Not through feeds. Through people.

And when that person finds this, they won't care that it's 3,000 words. They'll be grateful it's 3,000 words. Because finally someone said the thing that needed more than 47 seconds to say.

I'm not fighting the attention economy. I'm ignoring it. I'm writing for the small, intensely focused, absurdly high-value audience that is actively searching for depth in an ocean of noise.

And if this piece resonated - if you're still here, 3,000 words later - then you're one of them.

Welcome to the pond.

-parsa