.png)
August 23, 2026
First, a quick hello before the issue. You may have noticed the newsletter went quiet for a few weeks. That was on purpose. Behind the scenes, we moved this newsletter and all of our websites into a brand new home, built to serve you better. Same newsletter, same voice, better casa. If you are reading this, you made the move with me, and I am so glad you are here. Gracias. Now, let's get to work.
Back in May, I wrote about a number that stopped me in my tracks: 74% of AI's economic value was going to just 20% of companies.
I called it the 74/20 Rule. I told you about my mother's kitchen, how she cooked without recipes, and how everyone thought it was magic. It was not magic. Her pantry was stocked the same way every week. Her spices lived in the same spot. Her process never changed: prep the night before, start early, adjust as she went. The intuition everyone admired was sitting on top of a system nobody noticed.
That was the whole point of the 74/20 Rule. The people winning with AI do not have better tools than you. They have a system, and you have a collection of tabs.
So why am I bringing it back? Because I just spent my own summer living it. While this newsletter was quiet, my co-founder Nikki and I rebuilt every system behind our business: our websites, our email, our workflows, all of it redesigned with AI at the center instead of bolted on top. It was unglamorous work. It was also the single best proof I have ever had that the rule is real. The tools did not change. The system did. Everything got faster.
And while I was doing that, the data kept coming in. Three months later, the gap I warned you about did not close. It got wider.
No es lo mismo predicar que dar trigo. It is not the same to preach as to give wheat.
Talking about AI is still not the same as doing something meaningful with it. And the newest numbers show most people are still preaching. In other words: practice what you preach. That is the moral of this whole issue. It is the same thing Nikki and I tell every leader we work with at FlipWork: AI adoption does not spread through mandates or memos. It spreads when leaders become practitioners first.
Here is what has changed in the three months since I first wrote about the rule, and what has stubbornly stayed the same.
1. Everybody is using AI now. Almost nobody has been transformed by it. Gallup's latest workplace tracking shows that half of employed Americans now use AI in their role, with 13% using it daily. Usage keeps climbing every quarter. But here is the line that should stop you: only about 1 in 10 employees at AI-adopting organizations strongly agree that AI has actually transformed how work gets done. Usage is up. Transformation is flat. That is the 74/20 Rule playing out in real time.
2. The CEOs figured out the problem is not the technology. IBM's 2026 CEO study surveyed 2,000 chief executives this spring, and 83% said AI success depends more on people's adoption than on the technology itself. Yet only 25% of the workforce uses AI regularly in their jobs. The C-suite bought the tools. The work has not changed. Sound familiar?
3. Redesign still beats adoption, by a lot. The same IBM study found that organizations redesigning their core business areas around AI are four times more likely to deliver on their objectives. That echoes exactly what PwC found in April: the winners are twice as likely to have redesigned workflows around AI instead of layering it on top of old ones. Two different studies, one season apart, same answer.
The same beliefs I flagged three months ago are still holding smart professionals back. The evidence against them just got stronger.
Why we believe it: We signed up for the tool. We ask it questions. We feel productive. That must count for something.
Reframe: Three months ago, using AI at all still felt like an edge. Today half the workforce uses it. Usage is now table stakes, and Gallup's data shows it is not translating into transformation for 9 out of 10 people. The advantage has fully shifted from having AI to having a system around it.
What to do: Stop asking "Am I using AI?" and start asking "Have I changed how I work because of AI?" If the answer is no, the crowd you are standing in got a lot bigger since spring.
Why we believe it: Every AI pitch leads with efficiency. Faster emails. Quicker summaries. Less busywork.
Reframe: PwC found that the companies capturing the most value are focused on growth, reinvention, and new revenue, not time savings. The new IBM data adds a twist: 64% of CEOs are now comfortable making major strategic decisions based on AI-generated input. While most people use AI to shave minutes, leaders are using it to shape strategy. Time savings is the floor, not the ceiling.
What to do: The next time you sit down with AI, skip the quick task. Bring it a harder question: "What am I missing in my strategy?" or "Where could my work expand that I have not considered?"
Why we believe it: It feels responsible to wait, to learn more, to run another pilot before committing.
Reframe: Three months of new data confirmed the gap is compounding, not closing. The organizations that were already ahead spent the summer scaling. The ones that were waiting are still waiting, just further behind. Waiting is not caution. Waiting is how you fall behind while believing you are being strategic.
What to do: Pick one workflow this week. Not a side project, not a test. A real workflow you run every week. Redesign it with AI at the center. Run it for 30 days. Measure the difference. I just did this to my entire business, and I will never go back.
I am doing this section differently starting this week. One tool, explained properly, with a real activation step. Sampling three tools a week is exactly the tinkering trap this whole issue is about. Going deep on one is how you build a system.
The big AI labs all made the same move this summer: they stopped building chatbots and started building coworkers. The difference matters. A chatbot answers your question and forgets you. A coworker knows your context, works in your files, and finishes real deliverables.
Cowork is Claude's version, and it is the one I recommend you start with. You point it at a folder on your computer, it reads your context before you type a word, and it works directly in your actual files instead of handing you text to copy somewhere else. It is also the tool I used all summer to rebuild our business systems, so this is not a secondhand review. And the timing is right: as of this week, Cowork is available on mobile for every paid Claude plan, so the coworker now travels with you instead of living on your desktop.
Prompt to steal:
"I want to redesign how I [weekly workflow]. Read my context files first. Then walk me through the current steps I am likely taking, suggest which ones AI should handle, and flag the steps where my judgment is irreplaceable. Ask me questions before you build the new workflow."
Also on my radar: OpenAI just launched ChatGPT Work, which pulls context from your team tools to turn notes and drafts into finished work and is included in all paid plans, and xAI just launched Grok Bot in beta, persistent agents that sign in to your tools and work even when your laptop is closed, starting at $120 per seat for teams. Same shift, three flavors. I will go deep on each one here after I have put it through real work, because you deserve a practitioner's review, not a press release.
These are designed to take you from knowing to doing. Share them with the leaders in your circle.
1. Rerun the 74/20 self audit. If you did it the first time around, run it again and compare. Ask yourself honestly: am I using AI to save time, or am I using it to change how I work? Write down one workflow you could redesign this week. Not automate. Redesign.
2. Activate this week's tool. Thirty minutes, start to finish. Create a folder with three simple files: who you are, how you think, what you need. Open Claude, turn on Cowork, and point it at that folder. Then run one real task through it using the prompt above, on a workflow you actually do every week. Notice how different it feels when AI starts with your context instead of a blank page. That folder is the seed of your system.
3. Ask a growth question, not an efficiency question. Instead of asking AI to do something faster, ask it to help you think bigger. Try: "Based on what you know about my work, what adjacent opportunity am I not seeing?" The answer might surprise you.
The most valuable thing I learned rebuilding our systems this summer: do not start with the tool, start with the workflow. Map the workflow first, on paper, every step. Then ask AI to challenge the map before you automate anything. Automating a broken workflow just helps you do the wrong thing faster. Redesigning it first is what puts you in the 20%.
AI is still not the advantage. Your system is.
When I first wrote about the rule, that was a prediction backed by two studies. Today it is a pattern confirmed by every major dataset of the year, and by my own summer in the trenches.
The leaders pulling ahead are not the ones talking about AI. They are the ones practicing it. The wheat is on the table. It has been on the table for months. Practice what you preach. Stop preaching. Start planting.
♻️ Share this with a leader who is still stuck in theory mode.
¡Hasta la próxima, un abrazo fuerte! (Until next week, a big hug!)
