or, Twelve squirrels, one inbox: what happens when everyone can 10x their outputI have a theory, and it starts with squirrels. If you're wired like me — the kind of person who's been described (affectionately, mostly) as having twelve squirrels running around in their head at any given moment — the last two years have felt less like a technology shift and more like a jailbreak. For most of my career, the squirrels were a liability to be managed. You could only chase one at a time. The other eleven ideas — the training program, the tool, the workflow fix, the thing you noticed in the field that nobody's writing down — sat in a notebook, or a someday-list, or just evaporated. The bottleneck wasn't ideas. It was hands. One brain, two hands, twenty-four hours. AI changed the math. Now each squirrel can have its own project running. One agent is building a curriculum. Another is drafting the SOP from last week's transcripts. A third is modeling scenarios for a board conversation. A fourth is cleaning up a spreadsheet you've been avoiding for a month. You're not doing one thing at a time anymore — you're conducting. And for people whose minds were already parallel-processing, this isn't learning a new skill. It's finally having infrastructure that matches the architecture. That's my theory: the power users of AI aren't necessarily the most technical people — they're disproportionately the people who always had more threads than hands. The ADD brain, the neurodivergent brain, the "sorry, one more idea" brain. AI didn't give them more ideas. It gave every idea a body. I think that's genuinely beautiful. And I think it's about to cause a very real problem. The problem: output scaled. Attention didn't. Here's the asymmetry nobody wants to talk about: production just got an order of magnitude cheaper, but consumption didn't get one second faster. Every artifact I produce — every deck, doc, dashboard, and briefing — lands on a human. A GM, a colleague, a board member, a customer. And that human still has the same 24 hours they had in 2019. Most of them were maxed out before we handed the whole workforce a content firehose. The receivers didn't get an upgrade. Only the senders did. The numbers on this are staggering, and I want to be careful to use real ones, because the internet is full of fake statistics about fake content, which is almost too on-the-nose. On volume: In 2010, the world created about 2 zettabytes of data. By 2024, that figure had reached roughly 147 zettabytes — around 402 million terabytes every single day — and it's still compounding at over 20% a year (Statista, via Soax and Exploding Topics). For perspective, researchers have estimated that every word ever spoken by every human in history would fit in about 5 exabytes. We now generate that much data roughly every twenty minutes. Think about that against the full sweep of human history: from the invention of writing to the printing press was about 4,000 years. From the printing press to the internet, about 500. We've now compressed "more information than all prior human speech" into the time it takes to drink a cup of coffee. On authorship: In November 2024, something quietly historic happened. According to a study by the SEO firm Graphite, the number of new AI-generated articles published on the web surpassed the number of human-written ones for the first time. An Ahrefs analysis of 900,000 newly published pages found that about 74% contained AI-generated content — though interestingly, only about 2.5% were pure AI. The overwhelming majority are human-AI blends. Which sounds a lot like... all of us, doesn't it? On the "ghost internet": You may have heard the "dead internet theory" — the once-fringe idea that most of the internet is bots talking to bots. I'd always assumed the popular numbers were exaggerated, and some are. But the verified figure is wild enough: per the Imperva/Thales Bad Bot Report, automated traffic surpassed human traffic for the first time in 2024, at 51% of all web activity, rising to 53% in 2025. Cloudflare's network data points the same direction. Humans are now, measurably, the minority participant on the internet we built. A meaningful share of what's being produced at historic volume is being produced for, and consumed by, machines. The ghost internet isn't a conspiracy theory anymore. It's a line item in a security report. The signal problem
So here's where my optimism and my worry collide. I love that my twelve squirrels are all employed. I've shipped more useful work in the last eighteen months than in any comparable stretch of my career. But I've also started asking a harder question before I hit send: did I just create signal, or did I just create volume? Because the failure mode isn't that AI produces bad work. Increasingly, it produces good work — competent, polished, plausible work — at a rate no human review process was ever designed to absorb. When everything is well-formatted and articulate, polish stops being a signal of importance. The old heuristics ("this took someone real effort, so I should read it") are dead. And the person on the receiving end — the GM with fourteen techs and a truck down, the executive with 300 unread emails — doesn't experience your tenfold productivity as a gift. They experience it as weather. The scarce resource was never content. It was never even ideas. It's attention — and it just became the most oversubscribed asset in the working world. None of this should surprise us, honestly. The economist Herbert Simon called it back in 1971, warning that a wealth of information creates a poverty of attention. Gary Vaynerchuk built a whole career on the same insight a decade ago, arguing that we don't live in a content economy or an information economy — we live in an attention economy, and attention is the only asset that matters because it's the only one that can't be printed. Simon predicted it. Gary V popularized it. AI just proved them both right at industrial scale. And it forces a reframe that I think most of us haven't fully absorbed: your competition isn't the direct competitor to your idea. It's not the other proposal, the other vendor, the other training program. It's everything else competing for the same decision-maker's 24 hours — their inbox, their meetings, their kid's baseball schedule, the other eleven well-crafted, AI-polished artifacts that landed on their desk the same morning yours did. You're not competing on merit anymore. You're competing for minutes. And the minutes were spoken for before you showed up. What I'm trying to do about it I don't have a tidy framework yet (give me a week; one of the squirrels is on it). But a few working principles: Compress before you send. If AI helped you produce it 10x faster, use some of that saved time to make it 10x shorter. The kindest thing you can do for a receiver in 2026 is a ruthless executive summary. The artifact can sit behind it for whoever needs the depth. Decide what doesn't ship.* The discipline used to be finishing things. Now the discipline is killing things. Of my twelve squirrel projects, maybe three deserve another human's attention this week. The other nine can keep running quietly — or stop. Production is no longer proof of value. Make consumption part of the design. Before creating anything, I try to name the receiving human, their available minutes, and the single decision or action the artifact should produce. If I can't, the problem isn't the format. It's that I'm making content, not communication. Protect the human channel. The more the ambient internet fills with synthetic material and bot-to-bot traffic, the more valuable the verified-human moments become: the phone call, the ride-along, the tough conversation in good times. Those don't scale. That's exactly why they'll matter more. The bottom line AI handed the parallel-processors of the world — the twelve-squirrel people — the most powerful amplifier in the history of work. That's the good news, and I wouldn't trade it. But amplification without curation is just noise at higher volume. The winners of this era won't be the people who produce the most. They'll be the people who earn attention honestly in a world where attention is the last thing that can't be manufactured — the ones who treat their receiver's 24 hours as more precious than their own output.
0 Comments
One of the things people get most wrong about leadership is believing that training can replace accountability. It can't. You can send someone to every workshop on the calendar, and if nobody ever has the honest conversation about what's slipping, nothing changes. Accountability is the number-one overlooked tool in every leader's toolkit — and the strange part is that we don't train on it. Most of us were never taught how to do it. Plenty of leaders aren't even comfortable with the subject; it's treated as something negative, something to be hinted at or whispered about in the hallway. And too often, people hear "accountability" and assume it's just shorthand for firing someone. It's not. Done right, it's the most caring thing a leader does — and it almost never has to get that far. So instead of telling you, let me show you. Below is Chapter One of The Accountability Advantage, a short fable about a good leader, a good shop, and the ninety-second conversation that didn't happen. It's a quick afternoon read. |
| Here are a few of Eddie's top tips for becoming an industry leader:
|
these are just the thoughts of someone who used to know more but now knows less...
This website uses marketing and tracking technologies. Opting out of this will opt you out of all cookies, except for those needed to run the website. Note that some products may not work as well without tracking cookies.
Opt Out of Cookies
June 2026
February 2026
January 2026
September 2024
June 2023
April 2022
November 2021
February 2021
March 2020
February 2020
January 2020
October 2019
May 2019
January 2019
November 2018
September 2018
June 2018
March 2017
February 2017
October 2016
August 2016
July 2016
June 2016
May 2016
April 2016
March 2016
February 2016
January 2016
December 2015
November 2015
October 2015
September 2015