🫡 Hi, I’m Melvin. At Capital of One, we study the AI economy like an investor studies a market. But instead of positioning a portfolio, you position yourself: your work, skills, assets, relationships and attention. Every week, we bring you new ways to think, build and grow.

Everyone seems to have a list of skills you should learn next.

Most of these lists get it backward. Someone picks a bunch of skills they think are important. Writes a convincing argument for each. And conveniently, might have a course to sell you.

No thanks.

We followed the money through hiring data, changing business models, and corporate profits. Then worked backward to find which skills are gaining value.

One CEO was so convinced about where things were heading that he sacrificed $60 million in annual recurring revenue to get there first.

Let’s get into it.

Shift #1: AI Is Breaking the Career Ladder Before You Can Climb It

Companies increasingly need people who can judge AI’s work. But that judgment was often developed by doing the very tasks AI is now taking over. Where will the next generation of experienced employees come from?

The numbers:

  • Stanford (August 2026, payroll data): workers aged 22–25 in the most AI-exposed jobs sit about 19% below their expected employment trend.

  • Indeed (May 2026): entry-level job postings fell 7.5% year over year, while senior-level postings rose 14.7%.

  • ZipRecruiter (July 2026): 38% of employers shifted basic junior tasks to AI, while 31% raised experience requirements for entry-level roles because of AI.

The Accounting Firm Asking Juniors to Develop Judgment Earlier

Junior accountants at PwC used to spend years gathering data, checking numbers, and preparing audits. Now AI is handling more of that work, and PwC is changing how it trains new hires.

According to PwC’s Jenn Kosar, new accountants will spend less time preparing audits and more time reviewing AI’s work. They need to catch mistakes, question results, and understand what the numbers mean much earlier.

The Skill: Building Specialized Judgment

Your value depends on judging AI’s work: when to trust it, when to intervene, and what to do next. If you’re experienced, you have a head start worth sharpening. If starting out, you need to find new ways to build it.

HOW TO SHARPEN YOUR JUDGEMENT:

  1. Get unusually good at one small thing. A reader recently told me how much faster he improved after specializing. Makes sense. A generalist marketer might face the same problem three times a year. Someone focused on onboarding emails might face it three times a week. After a hundred attempts, they start spotting mistakes everyone else misses.

  2. Keep receipts on your decisions. That campaign you were certain would work? Six months later, you will have a dozen explanations for why it failed. Write down your reasoning before you know the outcome. Then compare what you expected with what actually happened.

  3. Steal the thinking, not the answer. Before watching a great designer fix a website, try spotting the problems yourself. Then watch what they notice, what you missed, and why they approached it differently. We’ve found you learn far more from those gaps than from watching someone do a great job.

Shift #2: The Money Is Moving From Deliverables to Results

Software companies used to charge per user. Agencies charged per hour. Now AI is putting pressure on both. Why pay for hours when the work takes minutes? Why pay for users when AI handles much of their workload?

More companies are starting to charge for results:

  • Salesforce charges $2 per resolved customer issue and nothing when its Help Agent escalates the issue to a human.

  • HubSpot charges $0.50 per resolved conversation and $1 per qualified lead recommended for outreach.

  • Sierra charges for completed outcomes and reached $100 million in annual recurring revenue in just seven quarters.

  • McKinsey says more than 30% of its fees are now tied directly to client outcomes.

The CEO Who Gave Up $60 Million in Revenue and Sold for $3.6 Billion

When Eoghan McCabe returned as Intercom’s CEO in 2022, growth had stalled. Then ChatGPT arrived, threatening to make Intercom’s traditional software obsolete. AI could now resolve customer problems without anyone needing to use it.

He shifted nearly 80% of R&D toward Fin, Intercom’s AI support agent. At the time, Fin generated less than 10% of the company’s revenue. He also overhauled pricing, charging $0.99 per resolved issue. These changes initially cost Intercom roughly $60 million in annual recurring revenue.

By March 2026, Intercom reached $400 million in annual recurring revenue, with Fin approaching $100 million. Three months later, Salesforce agreed to acquire the company for roughly $3.6 billion.

The Skill: Redesigning What You Sell

A web designer can promise a finished website. More sales? Different story.

The skill is figuring out what customers really want to achieve, how much of it you can deliver, and what they’re willing to pay for it.

HOW TO MAKE THE SHIFT:

  1. Look at what happens after you deliver. A prospect list still needs qualification. A website still needs to convert visitors. Find out what customers do with your work after receiving it. Taking responsibility for the next step could turn an ordinary service into something much more valuable.

  2. Get paid for results without betting everything on them. We’ve experimented with performance-based pricing and found that a base fee plus a performance bonus can work better than going all in. The base covers your costs. The bonus rewards the results you deliver. You get a piece of the upside without gambling your entire income.

  3. Watch who your pricing attracts. Performance-based deals can attract struggling businesses with little to lose. Sales Higher learned this firsthand and switched from pay-per-lead pricing to subscriptions. We would reserve performance-based deals for clients with proven demand. Otherwise, you risk financing someone elses business experiment.

Shift #3: The Cost of Testing a Business Idea Is Collapsing

Sam Altman recently made a great point: You can now test several ideas before betting months of your life on one. Launch a rough prototype, throw it in front of customers, and see if anyone bites. Nobody cares? Try the next one.

X avatar for @tbpn

TBPN@tbpn

.@sama says the advice to pick one thing to build and pour all your focus into it no longer applies. "The canonical advice that was correct two years ago was, 'Pick one, and put all your effort into it, because you can't build all 30 things.'" "You can now build all 30 things. …

9:36 PM · Sep 29, 2026 · 294K Views

70 Replies · 89 Reposts · 1.65K Likes

The Founder Who Booked 12 Events Before Building His App

Brian Shin and his girlfriend wanted to turn disposable cameras into an app for weddings and parties. They built a rough web version in a week or two and tried it at a friend’s Halloween party.

It kept breaking.

The guests loved it anyway.

Next, Brian wanted to see whether strangers would bite. So he set himself a target: ten event bookings before building the mobile app.

He found four through friends, then searched Instagram for people planning weddings and birthdays. He sent roughly 250–300 short messages. About fifteen people replied. Twelve booked.

That was enough.

Brian and his co-founder built the mobile app using AI tools, including Claude Code. In a May 2026 interview with Starter Story, he reported reaching $20,000 in monthly revenue within 83 days of launch.

The Skill: Demand Testing

Dating teaches a useful lesson here.

You can spend hours wondering whether someone likes you, analyzing every glance and asking five friends for advice. Or you can walk over and find out. The same goes for business ideas. Get better at walking over and finding out.

HOW TO STOP GUESSING:

  1. Test what you’re least sure about. We like starting with the assumption that could kill the idea. Brian took the same approach. He knew people enjoyed taking photos at parties. What he didn’t know was whether they would like his disposable-camera idea. You can save yourself months of work by figuring out which assumption you’re least confident about and testing that before building the rest.

  2. Find people who need it now. Brian found people planning actual weddings and birthdays. They had a deadline and a reason to care. People already spending money, using workarounds, or searching for solutions will give you a much better signal than people who simply like your idea.

  3. Make customers put something on the line. Compliments cost nothing. Signing up for a waitlist costs nothing. Getting someone to spend money, change their routine, or trust you with an important event takes more. If they won’t make a small commitment now, don’t bet months of work on their enthusiasm.

Shift 4: AI Is Creating a Customer Shortage

AI is squeezing businesses from two sides.

1) It’s making it easier for competitors to enter their markets. 2) AI assistants are changing how customers discover and buy products. They can compare options, recommend brands, buy products for customers.

Meaning? More competitors fighting for fewer chances to make their pitch.

Look at the numbers:

  • Lovable (June 2026): Users create roughly one million new projects every week. 80% of surveyed builders identify as non-technical.

  • Sensor Tower (July 2026): New App Store releases doubled to roughly 560,000 in the first half of 2026. Downloads grew just 2%.

  • Ahrefs (February 2026): The top-ranking Google result gets 58% fewer clicks on average when an AI Overview appears.

  • Salesforce (September 2026): Shopping journeys beginning with AI-powered search grew 200% year over year.

  • Shopify (October 2026): Shopify opened its catalog of more than one billion products to AI agents through Shop MCP. Agents can search products, build carts, and initiate checkout.

How Base44 Reached 250,000 Users Against Bigger Rivals

Maor Shlomo built Base44, an AI tool that lets anyone create apps without coding. He entered a market where Lovable and Bolt already had much bigger audiences.

On LinkedIn, Shlomo shared new features, revenue figures, and technical problems. Then users started posting the apps they built. He rewarded them with extra credits, giving people another reason to share their creations.

Within six months, Base44 reached ~250,000 users. In May 2025, Shlomo reported $189,000 in profit. A month later, Wix acquired the company for an initial $80 million.

The Skill: Creating Customer Preference

Notice that this is not distribution.

❯ Distribution gets your product in front of people.

❯ Customer preference makes them choose it over the alternatives.

Everyone keeps saying distribution will be the great moat of the AI era. We are not so convinced. Just while writing this article, Lovable announced that it can now create ads and run marketing campaigns.

So what is left? Giving customers a reason to seek you out, choose you over competitors, and recommend you to others. That's the skill of creating customer preference.

Maor Shlomo on Reddit.

HOW TO MAKE CUSTOMERS CHOOSE YOU:

  1. Find a reason to choose you over the market leader. Base44 couldn’t compete with Lovable’s reach, so Shlomo focused on customers who wanted everything built into one product. Find out what frustrates people about the leading alternatives. You need to be noticeably better at something a specific group of customers cares about.

  2. Become known before people need you. 6sense found that 95% of winning B2B vendors were already on the buyer’s initial shortlist. Find out where your potential customers learn, ask questions, and discover new products. Show up there consistently with something useful. When they’re ready to buy, you want to be a name they already know.

  3. Make your customers your next marketing channel. Shlomo noticed users sharing their Base44 creations and rewarded them with credits. Think about what your customers naturally want to show others. A designer shares a finished project. A runner shares a personal record. Make those moments bring you new customers.

Shift 5: AI Is Making Workers Faster, but Companies Aren’t Keeping Up

80% of employees in a McKinsey survey say AI makes them more productive. Only 37% report a positive impact on their company’s operating profit. Everyone is saving time. So why are companies not making more money?

  • Gartner (September 2026): 85% of business leaders plan to spend more on AI, but only 22% of organizations have successfully scaled it across multiple business units or adopted an AI-first approach.

  • Workday (January 2026): Nearly 40% of time saved with AI is lost correcting mistakes, rewriting, and checking its work.

  • McKinsey (July 2026): : Only 11% of companies have redesigned work around AI. Of those, 48% report meaningful results, versus 13% of companies mainly providing AI tools.

The AI Overhaul That Left Meta With More Problems to Fix

Zuckerberg wanted AI agents to make much smaller teams possible. In early 2026, Meta considered reducing some teams by as much as 60%.

AI helped employees produce far more code. Internal software changes increased 220% year over year. But changes delivering new or improved features to users rose just 36%. Meanwhile, major technical and security incidents increased 40%. Employees spent 70% more time fixing them.

According to an August 2026 Reuters investigation, Meta eventually abandoned plans for a second round of layoffs because of employee resistance and disappointing AI results.

The Skill: Finding the BottleNeck

When AI makes one part of a business faster, whatever comes next becomes more important. The skill is spotting where work starts piling up and figuring out how to redesign that part of the business.

HOW TO FIND THE WORK AI LEAVES BEHIND:

  1. Run the 10× test. We like asking: If AI made a task ten times faster, what would stop us from making ten times more money? A freelancer might build websites faster but lack clients. A logistics company might process orders faster but struggle with deliveries.

  2. Watch where work piles up. For a week, notice what keeps waiting. Drafts you haven’t reviewed? Customers you haven’t followed up with? Decisions you keep postponing? If you have a team, look at where their work gets stuck. If you work alone, your unfinished to-do list can tell you plenty.

  3. Always attack the constraint before automating. If you can produce ten proposals a day but still win only one client a month, making proposals faster won’t help much. Work on what limits the final result. Measure completed projects, sales, and profits instead of hours saved.

Shift #6: The Business of Bringing People Together Is Growing

Apparently, helping strangers making friends is becoming a pretty good business. As AI makes everything digital cheaper, we expect demand for real-world experiences to grow. A lot.

The numbers back it up:

  • Eventbrite (2026): 74% of surveyed 18–35-year-olds in the US and UK consider in-person experiences more important than digital ones. Another 79% plan to attend more events in 2026.

  • Live Nation (2025): Concert attendance reached approximately 159 million, up from 151 million in 2024. Concert revenue grew 10% to $20.9 billion.

  • ClassPass (2025): Fitness reservations increased 36% year over year, wellness reservations grew 37%, and bookings for community-based sports increased 24%.

Making Money by Getting People Off Their Phones

❯ The free running club that generated $640,000.

Owen Akhibi Herrera founded Peak and Pace in New York in 2024. The runs are free, but parties and trips cost money. In 2025, the community recorded over 30,000 attendances and generated more than $640,000 in revenue.

❯ The dinner organizer reporting €18 million in annual recurring revenue.

Maxime Barbier built Timeleft to organize dinners for strangers. Customers pay Timeleft for matching and reservations, then pay restaurants separately. By October 2025, it reported €18 million in annual recurring revenue and 150,000 monthly participants across 200+ cities.

❯ The founder earning revenue from hikes with other founders.

After moving to San Francisco, Alena Beliauskaya started organizing hikes and camping trips for founders. Her community, 12 Scrappy Founders, grew to over 600 members. By October 2026, it generated more than $10,000 monthly, with $2,000–$3,000 in profit from memberships, sponsorships, and events.

The Skill: Building Communities

Become good at creating experiences where strangers meet, form friendships, and want to keep coming back. You can build a business around this or bring customers together through one you already run. The relationships they form can give you a moat that competitors can’t copy.

HOW TO TURN STRANGERS INTO REGULARS:

  1. The activity does half the social work. Put twenty strangers in a room and tell them to network. Most will struggle to start a conversation. Take them hiking and suddenly they have something to talk about. A shared activity removes much of the awkwardness of meeting someone new.

  2. Find people whose social lives just changed. Someone who just moved to a city needs new friends. So do people changing careers, becoming parents, or starting businesses. The Mom Walk Collective grew from two mothers meeting for a walk to a community spanning over 650 cities. We would look for moments when people suddenly need relationships they don’t yet have.

  3. Bigger crowds can make your community worse. You organize an event for fifty people. Everyone has a great time. Next week, fifty different people show up. Nobody remembers anyone because nobody has met before. Bring the same twenty people together regularly and friendships start forming.

Shift #7: Capital Takes a Bigger Cut, and Workers Take a Smaller One

You can master every skill we covered and still spend your life building someone else’s wealth. You have to find a way to own a piece of the value you create.

  • BLS, Q2 2026: Workers received just 52.8% of US business output in wages and benefits, the lowest share since records began in 1947.

  • Goldman Sachs, September 2026: : Workers’ income share fell 7.5 percentage points since the 1990s, with 4.5 points linked to higher corporate profits, automation, and weaker bargaining power.

  • BEA, Q2 2026: US corporate profits reached an annualized $4.71 trillion, up from $4.33 trillion the previous quarter.

The Factory Workers Who Made $240,000 From the AI Boom

In 2023, KKR bought data-center cooling company CoolIT Systems and gave its employees ownership stakes. Then AI demand exploded, sending CoolIT’s value soaring. Three years later, KKR sold the business for $4.75 billion. Around 650 employees received payouts averaging $240,000. Some walked away with more than eight years of salary. All because they owned a piece of the business.

The Skill: Turning Your Work Into Assets

Start measuring your career by two things: what you earn and what you own. We've found it useful to ask what you will have accumulated after five years of doing the same work. More experience and a higher income? Or assets that can keep generating value?

❯ If you own a business: AI can dramatically increase your profit margins. But your competitors have access to the same tools. You want to build assets that grow stronger as you operate.

❯ If you don’t own a business: Your skills, experience, and relationships can help you earn a stake in a growing company or build something of your own. Find a way to participate in the value you help create.

HOW TO OWN MORE OF THE VALUE YOU CREATE:

1. Use your career to build something you own. Your industry knowledge and relationships can help you find problems worth solving. One reader spent a decade in construction before building SubBase, a material-ordering platform. His first customer was a subcontractor he knew.

2. Build a business that gets stronger as AI improves. Some businesses will become more profitable as AI cuts their costs. Others will lose value as AI makes their services cheaper. Before you build or buy, learn to tell the difference. We developed a five-year stress test to help you identify businesses that become more valuable as AI advances.

3. Buy a business and improve its economics. You don't have to start from zero to become an owner. Look for established businesses with paying customers but expensive, manual operations. AI can help you automate administration, improve sales, and serve more customers without adding staff.

Final Thought

The biggest mistake you could make is choosing what to learn based on what pays well today.

AI is changing what companies need, what customers will pay for, and where profits accumulate. Some skills will become cheaper. Others will become more valuable precisely because everything around them gets cheaper.

We spent this issue following those changes because that is where we believe the best opportunities are hiding.

Find the shift that gives you the biggest advantage, get exceptionally good at it, and build something you own along the way.

Five years from now, you will be glad you started.

—Melvin

At this point, we’re basically friends.

If you want to support Capital of One, there are three ways.

Become a member.

Spread the propaganda.

Share

Make some noise.

Leave a comment

Do all three and I may have to start paying you a referral fee.

Discussion about this post

Ready for more?