You Hired Einstein. He's Folding Laundry.
AI is the most powerful tool ever built. Why are you using it to fold laundry?
You Hired Einstein. He's Folding Laundry.
Imagine you hired the most brilliant mind in human history.
Not a consultant. Not a contractor. Einstein himself — the man who rewired how we understand physics, time, and the universe. You put him on the payroll, gave him full access to your organization, and then handed him a basket of laundry.
He folds it perfectly. Every crease sharp. Every shirt organized by color.
That's what most companies are doing with AI right now.
They have access to the most powerful reasoning and synthesis tools in human history, and they're using them to summarize meeting notes, clean up email copy, and generate first drafts of internal memos nobody reads.
Not because AI can't do more. Because nobody in the organization knows how to aim it at something that actually matters.
The Gap Isn't the Technology
McKinsey estimates that less than 30% of companies deploying AI have seen meaningful, measurable impact on their business results. Gartner says AI projects fail at a rate that would get any other technology initiative shut down. Forrester found that most organizations are running AI on processes that were already low-value before AI touched them.
This is not a technology problem. The tools are extraordinary. GPT-4, Claude, Gemini — these systems can synthesize thousands of documents, identify patterns across datasets no human team could process, reason through complex strategic problems, and generate outputs that compress weeks of work into hours.
The problem is organizational. Companies bought the capability without building the function that makes it work.
They hired Einstein and forgot to give him a problem worth solving.
What "Aimed Correctly" Actually Looks Like
The companies getting real returns from AI share one thing: a human in the loop who knows both sides of the equation.
Not the AI vendor. Not the IT department. Not a consultant who's never touched the tools outside of a demo. Someone who spent 20 years running the actual workflow — who knows where the data is dirty, where the edge cases live, where the model will confidently produce the wrong answer, and what "right" means in the context of this specific business at this specific moment.
That person doesn't just prompt the AI. They define what the AI is trying to solve. They evaluate whether the output is actually correct or just formatted to look correct. They translate what the model produces into a decision the organization will act on.
This is a leadership function, not a technical one. The technical work — the models, the APIs, the infrastructure — is increasingly commoditized. The judgment work — knowing which problem to point it at and whether it solved it — is not.
The companies running AI on laundry are missing this function. The companies getting 10x returns have figured out where to put it.
Why Smart Organizations Keep Getting This Wrong
This isn't a story about lazy companies or bad leadership. The people running these organizations are smart, experienced, and paying attention. They're getting this wrong for structural reasons.
First, AI adoption has been driven by vendors, not operators. The people selling AI tools are incentivized to make deployment look easy and outcomes look inevitable. They show the demo. They show the use case. They don't show the six months of failed prompts, dirty data, and organizational resistance that precede any real result.
Second, most organizations built their AI strategy around efficiency, not capability. The business case was cost reduction: automate the repetitive work, reduce headcount, improve margins. That framing pushes AI toward the lowest-value tasks in the organization — exactly the laundry pile. You can measure the savings easily. You can't measure what you never tried.
Third, and most important: companies don't have a job description for the person who closes this gap. They have AI engineers who build the systems. They have business analysts who use the outputs. But the role that sits between them — the person who understands both the domain and the tools well enough to connect them — doesn't exist on most org charts yet.
That vacancy is the problem.
The Role Nobody Has Written the Job Description For
There's a function emerging at the intersection of domain expertise and AI fluency. It doesn't have a standardized title yet. Some companies are calling it AI Strategy Lead. Others are calling it Head of AI Operations, or AI Transformation Director, or just burying it in a VP title and hoping it gets figured out.
What the role actually requires: 15-20 years of deep domain knowledge in a specific function — finance, supply chain, sales, product, operations — combined with enough working fluency in AI tools to know what they can and can't do, and the organizational credibility to translate between them at the executive level.
This is not the AI expert who optimizes the model. It's not the domain expert who uses AI as a productivity tool. It's the person who sits at the intersection and makes the whole thing work.
Every organization deploying serious AI investment right now either has this person or is about to realize they need one. The companies that have it are the ones showing up in the McKinsey 30%. The ones that don't are the ones with Einstein folding laundry.
The window to position yourself in this role — whether inside a company or as an outside resource — is open right now. Not for long.
The Numbers That Make This Concrete
The average enterprise is now spending over $85,000 per month on AI tools and infrastructure. That number is up 36% year over year. Token costs fall 10x per year, which means as AI gets cheaper, companies buy more of it. The spend accelerates even as the per-unit cost drops.
What they're not buying — yet — is the capability to convert that spend into outcomes.
There is a measurable gap between what companies invest in AI and what they actually capture from it. Analysts are calling it the AI productivity paradox: more investment, more tools, more deployment, but the productivity gains remain concentrated in a small percentage of organizations.
That gap is not going to close with better models. The models are already extraordinary. The gap closes when the right human is in the right seat, pointed at the right problem.
That human is worth a significant premium right now precisely because most organizations don't have one and are starting to realize it.
Two Paths. Both Built on What You Already Know.
If you've spent 20 years in a functional domain — supply chain, enterprise sales, financial operations, product development, healthcare administration — you are closer to this role than you think. You already have the domain expertise. The AI fluency layer is learnable. Not in years. In months, with the right framework.
The W-2 track: You become the person your organization cannot restructure away. Not by managing a bigger team or owning a larger budget, but by being the leader who turned $85K a month in AI spend into $3M in measurable outcomes. That's a different conversation in a performance review. It's a different profile in a reorg. It's the difference between being the person who survives the next round of cuts and the person who decides who stays.
The professionals building this capability now are not waiting for their company to hand them a roadmap. They're showing up with one. They're identifying the high-value problems in their domain, deploying AI against them with discipline, and documenting the results. By the time the job description gets written, they're already doing the job.
The fractional track: Four to five companies at $50K–$75K each. $200K–$350K annually. Twenty to thirty hours a week. You bring 20 years of domain expertise plus AI fluency to problems that Google won't touch because they're too specific, McKinsey won't touch because the market's too small, and the companies facing them can't solve internally because they don't have anyone who speaks both languages yet.
This is not a consulting practice built on slide decks and frameworks. It's built on the ability to walk into an organization, look at how they're deploying AI, identify the highest-value application they're missing, and get it running. That capability commands a premium. The market for it is growing faster than the supply of people who can deliver it.
In both cases, the foundation is the same: domain expertise you already have, plus a layer of AI fluency you can develop, plus the judgment to connect them. That combination is rare right now. It won't always be.
The Laundry Isn't the Problem
Einstein folding laundry isn't a tragedy because of the laundry. It's a tragedy because of everything that didn't get solved while he was folding it.
The companies running AI on their lowest-value work aren't just leaving money on the table. They're building organizational habits that will be hard to reverse. Teams that treat AI as a productivity tool rather than a strategic capability. Leadership that evaluates AI by whether it saves time on tasks rather than whether it changes what's possible. A culture that never discovers what the tools can actually do when pointed at a real problem.
The companies that get this right in the next 18 months will not just have a productivity advantage. They'll have a strategic advantage that compounds. Every high-value problem they solve with AI teaches them how to solve the next one. Every outcome they capture gives them data and organizational confidence that the laundry-folders won't have.
That's the fork. Not AI versus no AI. Aimed correctly versus aimed at the laundry.
Which side are you on?
Ready to Figure Out Where You Fit in This?
Written by
Bill Heilmann