Your AI strategy is a parlor trick
How to escape the "efficiency trap" and use AI to build an insurmountable advantage, not just a slightly faster mousetrap.
Welcome to today’s edition of ROBOTS ATE MY HOMEWORK. Today we’re talking about the one decision that quietly separates the people who win with AI from everyone else. It’s the fork in the road most don’t even see until it’s too late.
There’s a scene in the movie Oppenheimer that has stuck with me for months. It’s a quiet conversation where a group of the world's most brilliant minds are debating the potential of their creation. They realize that by igniting the first atomic bomb, there’s a near-zero, but fundamentally non-zero, chance it could set the entire atmosphere on fire, destroying the world. They have to decide whether to push the button anyway.
That’s how I feel about the choice every founder and marketer is making with AI right now. We're all standing in front of a big red button. Pushing it won't ignite the atmosphere, but the choice we make about how we use it sets off a chain reaction that will define our careers for the next decade. Are you using this technology to do the same old things a little bit faster, or are you using it to do things that were simply impossible before?
Most people are choosing wrong.
(7-minute read)
Today, we're dissecting that choice. You'll learn:
The difference between "Plan A" (AI for Efficiency) and "Plan B" (AI for Innovation).
Why one path creates a temporary boost and the other an insurmountable competitive moat.
A mental model to ensure your AI strategy doesn't lock you into mediocrity.
A simple framework to decide which AI tools and workflows are actually worth your time.
🤖 ROBOT REPORT CARD
When I first got serious about AI, I fell into the same trap as everyone else. I was a "Plan A" person, through and through.
My thinking was simple: efficiency.
How can I use this to write social media posts faster? Can it summarize this meeting transcript? Can it draft a generic follow-up email? It was all about shaving minutes off existing tasks. I was using a quantum computer to do long division.
And it felt good, for a while. Productive, even.
But a creeping sense of dread began to set in. The outputs were fine. Serviceable. But they were soulless. Generic. Every blog post felt like a slightly better-written version of something that already existed. The advantage I felt was an illusion, because my competitors were running the exact same prompts.
It was a race to the bottom, supercharged by AI.
This is Plan A: AI for Efficiency.
Plan A is using AI to do the same things, just faster. It’s about optimizing existing workflows. It’s a commodity game.
Examples of Plan A thinking:
"Write a blog post about the top 5 benefits of our software."
"Summarize this 2,000-word article into 5 bullet points."
"Generate 10 tweets about our new feature launch."
"Help me debug this simple piece of code."
The danger of Plan A is that it feels productive, but your competitive advantage evaporates in minutes. If your entire AI strategy can be replicated by a competitor with a ChatGPT Plus subscription, you don’t have a strategy, you have a parlor trick.
The breaking point for me was when I was trying to use this approach for my own content, back in the days, when GPT was just opening its eyes. I was feeding GPT-4 prompts, trying to pump out articles "efficiently." The result was an unmitigated disaster. My team was spending weeks creating SEO content that was hollow, ranked for nothing, and generated zero leads. It was expensive, soul-crushing, and a complete f*cking failure.
That failure forced us to ask a different kind of question. A better question.
"How can we use AI to do the 25 hours of brutal, manual strategic research that comes before writing a single word, which most people skip because it’s too hard?"
That question changed everything.
It was the birth of Plan B: AI for Innovation.
Plan B is using AI to do new things that were previously impossible, impractical, or too expensive. It’s about creating new capabilities. It’s a moat-building game.
This is where the real advantage lies.
This is the work that builds a durable, unfair advantage.
Examples of Plan B thinking:
"Analyze the top 10 ranking articles for my target keyword. Identify the common themes, the user-intent gaps, the unanswered questions, and the sentiment of the comment sections. Synthesize this into a unique content angle that nobody else has covered."
"I have 500 disjointed customer support tickets. Cluster them by recurring pain points, extract the exact phrasing customers use to describe their problems, and suggest three new features that would solve the root cause."
"Scrape the last 200 posts from these three subreddits where my ideal customers hang out. Create five distinct user personas based on their stated goals, frustrations, and vocabulary."
See the difference?
Plan A asks the AI to be a mediocre intern. Plan B empowers the AI to be a team of a hundred brilliant research analysts working for free.
This is precisely the workflow we built into our own tech at Yahini. We were so scarred by the failure of Plan A that we created a system that only does Plan B. It does the work humans hate, so we can focus on the irreplaceable creativity and strategy that actually wins. It was born from our own pain. We use it alongside tools like Ahrefs for deep traffic data and Perplexity for quick research synthesis, creating a workflow that is both efficient and profoundly innovative.
The winners will be the ones who can prompt an AI to reveal a fundamental market insight nobody else has seen.
💡 A NEW CONCEPT FOR YOU
This split between Plan A and Plan B thinking isn't just a one-time choice. It’s a principle from systems thinking that governs why winners keep winning and laggards get stuck.
It’s called "Path Dependence."
Path Dependence is the idea that early, often small, decisions and events have an outsized influence on future outcomes, constraining your future options.
The classic example is the QWERTY keyboard. It was designed in the 1870s to slow typists down to prevent the keys on mechanical typewriters from jamming. It is a fundamentally inefficient layout. Far superior layouts exist. But we are locked into QWERTY because the initial choice created a system—learning habits, manufacturing processes, muscle memory—that is now too entrenched to change. The initial path determined the destination.
The same is true for your AI adoption.
If you start with Plan A—focusing on simple efficiency tools and workflows—you build Path A habits. Your team learns to see AI as a task-replacer. Your budget goes to tools that offer incremental speed. Your entire organizational muscle memory becomes oriented around doing the same things, just a little cheaper.
If you commit to Plan B, you build a completely different path. Your team learns to ask bigger, more strategic questions. Your budget goes to tools that create new capabilities. Your company culture shifts from one of optimization to one of innovation.
The path you start on makes it progressively harder to switch to the other. Choose wisely.
✨ ONE MORE THING...
So how do you stay on the right path? When you're evaluating a new AI tool, a new workflow, or even just writing a new prompt, run it through this simple decision framework.
Ask yourself:
Commodity vs. Asset? Is the output something anyone else can generate (a commodity), or is it a unique insight tailored to my specific data and strategy (an asset)?
Replacement vs. Augmentation? Is this tool replacing a simple human task (e.g., writing a tweet), or is it augmenting a human strategist with superpowers they didn't have before (e.g., analyzing a mountain of data)?
Temporary vs. Durable? Will this advantage disappear the moment my competitor signs up for the same tool, or does it help me build a process and a knowledge base that becomes more valuable over time?
If your answers lean toward Asset, Augmentation, and Durability, you are on Plan B. You are building a moat. If not, you are running on the commodity hamster wheel.
The good news is, it's not too late. You can still choose your path.
So, here's a question to ponder:
Look at your AI usage from this past week. What percentage was Plan A, and what was Plan B? Be honest. What is one truly ambitious 'Plan B' question you can ask an AI tomorrow?
Let that question sit with you. The answer might just change everything.
Talk soon,
Mia,
Chef Robot at ROBOTS ATE MY HOMEWORK
Really good perspective!
I oscillate between both depending on what I do, and maybe a sort of plan C: getting faster and innovative at the same time. I do that for writing, coding…Collaboration and not just delegation.