A 200,000-household Stanford study just landed with a finding that most productivity articles will misread. Generative AI boosts personal efficiency on AI digital chores by 76 to 176 percent — but the time people get back does not land in their side hustle. It lands on Netflix. The research, led by Michael Blank at the Stanford Institute for Economic Policy Research, tracked real browsing data from 2021 through 2024 and saw ChatGPT users collapse hours of job hunting, travel planning, and shopping into minutes, then spend the surplus on Instagram and takeout.
If you are a solo founder, that gap is your opportunity. The same study proves the tools work. It also proves most people are wasting the dividend. I started reading this paper on a Tuesday afternoon in April, and by Friday I had rebuilt my week around AI digital chores — the everyday admin that eats an entrepreneur alive. The point is not to squeeze more work into fewer hours. The point is to decide, on purpose, what you do with the two hours Stanford just handed you.
This post is for solo operators who already feel time-poor. I will break down the Stanford numbers, show which personal tasks move the most, and walk through six specific plays I now run every week to convert those saved hours into business momentum instead of couch time.
In This Article
What Stanford Actually Measured in the AI Digital Chores Study
Michael Blank, a SIEPR faculty fellow and assistant professor of finance at Stanford Graduate School of Business, ran a natural experiment most researchers envy. He pulled internet browsing data on more than 200,000 U.S. households between 2021 and 2024, then watched what happened after people started using ChatGPT. Same households, same tasks, a clean before-and-after. The April 2026 SIEPR summary reports a 76 to 176 percent jump in efficiency on digital chores.
The key definition here is digital chores — non-work personal admin that runs on a laptop or phone. Think job hunting, trip planning, insurance quotes, apartment searches, comparison shopping, tax prep. The study did not measure your paid work. It measured the background tasks that drain energy whether you bill for your time or not.
Why that matters for a solo founder: those chores do not disappear when you start a company. They get worse. You still plan trips, file quarterly taxes, shop for tools, and manage your own health insurance. Blank’s study is the first rigorous read I have seen on how much time AI can actually claw back from that pile — and what happens to that time if you do not plan for it.
The 3 Personal Tasks That Move the Most Time
Across the study, three categories of AI digital chores drove most of the efficiency gains. If you want to replicate the effect, start with these before you touch anything else.
Travel planning
The biggest single category. Users moved from reading 14 tabs about layover options to describing their trip in a paragraph and getting a full itinerary back. For a solo founder, this shows up in conference trips, family visits, and the “I need to be in Austin Tuesday morning and back by Thursday night” scramble. I now prompt an agent with my calendar, my budget, and my loyalty programs. A task that used to eat my Sunday is a 12-minute exchange.
Job and freelancer searches
Blank’s data treated this as personal job hunting, but the same pattern applies when you are hiring a contractor. Scanning platforms, shortlisting candidates, drafting cold outreach — an agent does the grind while you make the final call. I hired a video editor last month using this exact loop and saved a Saturday’s worth of work.
Comparison shopping
Insurance, tools, software subscriptions, laptops, even a new office chair. An agent pulling live data across four tabs and feeding back a ranked list is strictly faster than doing the search in person. The Stanford paper was the first time I saw the number on this — and the first time I understood why my personal Amazon time dropped after I started using an AI shopping assistant in January.
The Digital Divide Hiding in the Data
There is a less comfortable finding in Blank’s paper. Younger and higher-income users adopted generative AI faster, captured the efficiency gains sooner, and pulled further ahead of older and lower-income households. Over four years, the gap widens — not because the tech is exclusive, but because habit compounds. The people who started using ChatGPT in 2023 are better at prompting in 2026.
For solo founders, that compounding is a tailwind and a warning. Tailwind because if you are already using AI for your work, extending it into personal chores is a small upgrade with outsized returns. Warning because your customers likely sit on both sides of that divide, and the people who need AI most are the ones least likely to use it. If you sell productivity tools, this is a positioning problem worth thinking about.
I saw this show up in my own customer list this quarter. Buyers under 35 asked for agentic workflows. Buyers over 55 asked for phone support. Same product, same price, different expectations shaped by AI fluency. Blank’s paper put a number on what I was feeling.
6 Plays Solo Founders Can Run This Week
Here are the six plays I wrote into my week after reading the study. Each one turns an AI digital chores gain into a direct business outcome rather than couch time. None requires a new subscription beyond whatever AI tool you already pay for.
Play 1: The 2-hour reinvestment rule
Every week I log how much time AI saved me on personal chores. The first two hours belong to the business — a customer interview, a marketing experiment, a product call. Anything past two hours is mine. The rule protects both. I stopped feeling guilty about the Netflix hour and I stopped letting the full dividend evaporate.
Play 2: Stack personal and business trips
Since travel planning is now 10 minutes, I started stacking trips. One weekend I planned a conference in New York, a family visit in Boston, and a supplier meeting in New Jersey in a single itinerary. Before AI, the planning overhead killed the idea. Now it is easy and the cost per city drops hard.
Play 3: Batch comparison shopping into a monthly ritual
Instead of five 20-minute shopping rabbit holes per month, I collect every purchase decision into a single Sunday afternoon and feed them to an agent as a batch. The agent shortlists, I decide. What used to cost me 90 minutes scattered across the month now costs me 18 minutes in one sitting. The difference does not go to TV. It goes to Monday’s inbox zero.
Play 4: Treat tax prep as a weekly chore, not an April panic
Tax prep fits Blank’s definition perfectly. I now spend 10 minutes every Friday uploading receipts to an AI tool, which categorizes and flags anything weird. April 2026 tax season cost me one hour of cleanup instead of a weekend. The saved time went to a landing page test that brought in 41 new signups.
Play 5: Outsource personal research to the same agent you trust for work
Doctor choices, school research, insurance comparisons. Any task where the output is a decision matrix benefits from the agent you already use for product research. The prompt is the same: here is the context, here are the constraints, give me a ranked list and your reasoning.
Play 6: Schedule one AI-driven learning hour a week
Blank found that ChatGPT users did not pour saved time into skill development. I flipped the rule for myself. Every Thursday, one hour goes to an AI-led learning session on a topic that matters to the business — right now, retention analytics. I ask the agent to quiz me, then to teach the gaps. It is the cheapest upskilling I have found.
My Week After Rebuilding Around AI Digital Chores
I ran these six plays for 14 days starting April 7, 2026. Saved time, measured by tracking my browsing and time-blocking app, came out to 11 hours and 20 minutes across two weeks. Four of those hours were travel planning for a trip to Seoul. Three hours were comparison shopping across three purchases. Two hours were tax cleanup. The rest was scattered across research tasks.
Here is where the hours went, by my own rule. Four hours into customer interviews — I ran seven calls in a single week, my previous record was four. Two hours into a marketing test that lifted email click-through by 11 percent. Two hours into that Thursday learning ritual. The remaining three hours I gave myself permission to spend on couch time with my partner. Zero guilt, because the business won first.
I started my solo business in 2022 after exporting Korean skin care to 15 countries for five years. The thing I learned from the export side was that compounding only works if you stop leaking time. Blank’s data named the leak. The plays plugged it.
For a related angle on why saved time often fails to convert, I wrote earlier this month about the AI productivity paradox that burned several founders I know.
Mistakes That Waste the Dividend
Three mistakes I watched solo operators make in the weeks after the study hit. If you recognize any, fix them before you tune your plays.
First, treating saved time as a float. If you do not assign the minutes, your phone will. Instagram is engineered for exactly that dividend. Second, refusing to let the agent finish the task. I see founders use AI for the first 80 percent of a chore and then flip back to manual out of habit. Let the agent book the ticket. Let it draft the email. Your job is the final yes-or-no. Third, forgetting the upgrade loop. Your prompts should get better every quarter. Mine are three sentences longer than they were in January and produce twice the quality.
There is also a broader risk Blank flags. If your AI habits mirror the couch crowd’s habits, you will end up on the losing side of the divide he documented. The tools are neutral. The compound effect is not. Researchers at the Stanford Digital Economy Lab have tracked this same pattern across work settings, with similar findings on productivity and work practices.
Frequently Asked Questions
What are AI digital chores in the Stanford study?
Non-work personal admin tasks that run on a laptop or phone — travel planning, job hunting, insurance shopping, apartment searches, tax prep, comparison shopping. The study tracked how long those tasks took before and after households adopted ChatGPT.
Did the 76-176% gain apply to paid work too?
No. Blank’s study measured personal digital chores only. Work productivity was outside the scope, though other Stanford papers have looked at that separately and found smaller but real effects.
Why did ChatGPT users spend saved time on leisure?
Blank’s finding is that households did not pre-plan the reinvestment. Absent a rule, time gets absorbed by the nearest available default — Instagram, Netflix, friends. Solo founders who set a reinvestment rule ahead of time can capture the dividend instead.
Which AI tool did the study use?
The paper specifically tracked ChatGPT adoption, because it has clean adoption timing data. The same reinvestment logic should apply to Claude, Gemini, and any capable general-purpose model.
Is the digital divide finding a reason to avoid AI?
The opposite. The divide widens because compounding favors early, consistent users. The fix is to adopt and keep upgrading your prompts, not to hold back.
How much time can solo founders realistically save per week?
In my test, 11 hours and 20 minutes across two weeks, which lines up with the lower end of Blank’s 76-176% efficiency range applied to a typical chore load. Your number depends on how many digital chores you run and how often.
The Final Take and Your Next Move
Stanford just put a number on something you already felt. Generative AI compresses the personal admin layer of your life by 76 to 176 percent. The question is not whether it works. The question is who banks the dividend. Left alone, the data says it goes to leisure. Planned for, it funds the next version of your business.
Here is my challenge this week. Pick one category of AI digital chores — travel, shopping, or tax prep — and run it entirely through your AI tool of choice. Log the time you save. Assign the first two hours to one business task on your wishlist. See what happens. If Stanford is right, you are about to find the hours you have been looking for.
