The clip everyone is sharing

David Fear’s Rolling Stone interview “I Didn’t Want to Be the Angry Guy” appeared on September 22, 2026. In clips now going viral, Burr argues that AI should replace highly paid chief executives instead of rank-and-file workers. No text transcript verifies his exact wording, so treat the clip as a paraphrase. The thrust is unmistakable: the corner office should be first in line for automation, not last.

It landed because it inverts the usual script. For a decade, the automation debate has centred on cashiers, drivers, warehouse staff and, lately, junior coders. Everyone assumed the safe jobs were the ones requiring thought. Strip out the comedy and the point is economic. A machine doing expensive cognitive work for a fraction of the price raises an awkward question. Why does the priciest employee in the building stay exempt?

Can AI Replace CEOs? The Case For

Take the joke out and the logic is simple. A chief executive is, in large part, an information processor: market data goes in, capital allocation decisions come out. Machines now do that faster, more consistently and more cheaply than any human. If the job is mostly pattern recognition and resource allocation, the maths favours automating the CEO. The salary at stake dwarfs a warehouse worker’s.

The question for boards is therefore not whether the idea is outrageous. It is whether the numbers justify it.

The math: $22.8 million vs the median paycheck

The AFL-CIO 2026 Executive Paywatch report puts the number at $22.8 million. That was the average 2025 pay for S&P 500 chief executives, excluding Elon Musk. That is 312 times the median US worker, up from 285 times in 2024. Read that again: the ratio widened in a single year.

Put it another way. One CEO’s annual pay could fund hundreds of median salaries. It could also bankroll a serious programme of AI investment. Replacing even a fraction of that compensation with software is worth millions a year. Automation has always gone after expensive labour first. It has simply never gone this high up the org chart before.

It already happened: the companies that gave AI the corner office

But can AI replace CEOs in practice? Two companies decided to find out back in 2022.

In August 2022, NetDragon Websoft made the appointment. The Hong Kong-listed firm (stock 777) named Ms. Tang Yu as Rotating CEO of subsidiary Fujian NetDragon Websoft. Tang Yu is an AI-powered virtual humanoid. Chairman Dejian Liu said: “We believe AI is the future of corporate management, and our appointment of Ms. Tang Yu represents our commitment to truly embrace the use of AI to transform the way we operate our business, and ultimately drive our future strategic growth.” Her brief was concrete: streamline process flow, act as a real-time data hub, improve risk management and support talent development.

Days later, Polish luxury rum maker Dictador hired Mika, a humanoid robot built by Hanson Robotics, as chief executive. Dictador signed the contract on August 30, 2022, and her “career” launched on September 1. Mika joined the board, taking responsibility for the Arthouse Spirits DAO project and communications with the DAO community. Europe president Marek Szoldrowski called the decision “revolutionary and bold.” The appointment drew wide coverage, including a detailed report by FoodBev.

A caveat: both moves were at least partly publicity. Nothing verifies that either Tang Yu or Mika still holds the role in 2026. Treat them as experiments, not precedents. But they proved the barrier was never technical. An AI can already sit in the org chart. The barrier was cultural, and culture is shifting.

Even CEOs agree: half say AI could do their job

The most surprising data comes from chief executives themselves. A 2023 survey by edX and Workplace Intelligence polled 800 executives, including more than 500 US CEOs. The result: 49% believe AI should automate or replace most or all of their role. In contrast, only 20% of workers said the same. Nearly half of sitting CEOs think a machine could do their job.

Beautiful.ai surveyed 3,000 US managers in April 2026. It found that 55% say AI could reduce their own pay. Another 58% say AI output matches or beats an experienced manager. And 35% believe replacing employees with AI would benefit their company, up from 23%. The people closest to the technology are the least sentimental about it.

Gartner polled 469 CEOs and senior executives in April 2026. It found that 80% expect AI to force a high-to-medium degree of change in operational capabilities. Investors see it too. In our interview with internet investor Toine Rodenburg, the theme was how quickly AI is reshaping company building.

Not everyone is convinced. Nvidia chief executive Jensen Huang spoke at the AI Summit in Mumbai in October 2024. He said AI can handle 20 to 50 percent of a given job. His caveat: “in no job can [AI] do all of it.” Asked if AI could replace him, he said: “absolutely not.”

What AI does better than a human CEO

An AI chief executive never sleeps, never forgets a KPI and can model thousands of scenarios before breakfast. Tang Yu’s job description is the template. A real-time data hub never misses a risk signal. It tracks operations and talent metrics around the clock, with no human calendar slowing it down.

Then the cost curve. CEO pay climbs toward 312 times the median worker. Meanwhile, open-source models are slashing the cost of AI capabilities. Machine intelligence gets cheaper every year; corner-office intelligence gets dearer. The curves point one way.

And consistency: software has no moods, builds no empires and never falls in love with its own strategy. Where decisions are genuinely data-driven, the absence of ego is a feature.

What AI cannot do: accountability, judgment, and trust

This is where the robot-CEO case breaks down. Scott Hartkopf of Threshold/AI put it sharply in February 2026. A board can fire a human chief executive, but “what happens when the CEO is software?” After all, someone must answer for decisions: shareholders, regulators and courts all need a human being.

Patrick Kamba made the cultural version of the point in CEOWORLD in May 2026. “Employees notice accountability” when leaders over-automate. Teams start asking where judgment really resides. A workforce unsure who is deciding will not follow with conviction.

Satish Mekerira offered the philosophical version in July 2026. “Algorithms cannot possess moral agency… AI can generate the map, but only a human can bear the weight of the journey.” Gregory Downey added the same month. “AI does not empathize. It predicts language associated with empathy… because it cannot be moved, it cannot truly move another human being.”

The through line: leadership is not just decision-making. It is owning decisions, absorbing uncertainty and carrying people through change. Software cannot do that, because when things go wrong, someone always has to answer.

The realistic future: the augmented executive

So can AI replace CEOs? The math says maybe, but accountability says not yet. The realistic outcome is a smaller, cheaper, AI-augmented C-suite. Fewer executives will wield machine intelligence. Humans keep judgment for the calls carrying moral and legal weight.

The pressure will hit pay first. Fifty-five percent of managers expect AI to compress their compensation. Executive pay cannot sit at 312 times the median worker forever. Boards once justified eight-figure packages as the price of rare talent. Now they must explain why they cannot rent that talent, in part, from a data centre.

For operators, the lesson is practical. The executives who thrive will treat AI as leverage, not threat: machines handle information, humans handle responsibility. Similarly, smaller firms face the same squeeze. Our interview with Chirag Shah, founder and CEO of Nucleus Commercial Finance, explores the pressure. Our entrepreneurship and startups hub tracks how company building is changing.

Burr’s joke will keep circulating because it holds a truth every pay ratio sharpens. The automation debate was never about whether machines can do the work. It is about whose work machines automate first. For a century, the answer was whoever cost least to keep. The next decade may finally invert that.

The experiment is already running. Every board that benchmarks executive pay against machine capability is, knowingly or not, voting on Burr’s proposition.

The post Should AI Replace the CEO Instead of Workers? The Math Says Maybe appeared first on CEO Medium.

I searched the CEO Medium interview archive for founders who talked openly about pricing. I wanted to know what they charged at launch, what they changed, and what revenue did next. Their stories line up with the latest industry data on pricing. Below are the seven pricing mistakes first time founders make most. Each one comes with the numbers behind it and a fix.

The pricing mistakes first time founders make most often

Each mistake below comes from a real founder interview or a published industry study. None of them require a finance degree to fix.

Mistake 1: You charge too little because your confidence is low

Lisa Collum founded Top Score Writing in 2011. Her classroom writing lessons lifted her students’ state test pass rate from 38% to 95%. Over the next two years it reached 100%. She went on to work with more than 1,000 schools across the country. Her results were elite. Her prices, however, were not. As she told CEO Medium: “At first, I often second-guessed myself. I felt nervous meeting schools and selling my products. I hesitated to charge much because I lacked confidence. Over time, the results gave me proof. I now feel ready to take both businesses to the next level.”

Of all the pricing mistakes first time founders make, this is the most common. It has nothing to do with the market. Founders price their own insecurity instead of their value. Still, the fix is blunt. Write down your three strongest customer results, then set a price that matches them. If your work moves a pass rate from 38% to 100%, you are not the budget option. Stop introducing yourself as one. Confidence follows evidence. Collect the evidence first and let the price follow.

Mistake 2: You set a price once and never revisit it

Jeremy Finlay, creator of the Salesumentary video method, learned this firsthand. His CEO Medium interview puts it plainly: “Pricing was a big obstacle for the team early on. Many small businesses have the tendency to underprice their work and undervalue their time. It took Jeremy a few clients to realize he was charging far too little for his services.” An old mentor insisted that he raise his prices. So he did, and then he raised them again. Still, demand held firm. His rule for clients today: “We don’t work with folks who are brand-new anymore. We work with those who are the best at what they do.”

Finlay’s instinct matches the data. Businesses that update pricing at least every six months see nearly double the ARPU gain of slower peers. That is the finding of Paddle’s ProfitWell Report, which studied roughly 5,000 subscription companies. Pricing is not a one-time decision. Treat it as a habit instead. Put a pricing review on the calendar twice a year.

Mistake 3: Nobody on the team owns pricing

Here is an uncomfortable number. An OpenView survey of more than 1,000 SaaS executives found a blind spot. 55% of expansion-stage SaaS companies ($1M to $20M ARR) had nobody whose job included pricing. More than half let pricing drift with no owner. It is also one of the quietest pricing mistakes first time founders make. The damage, however, stays invisible until you look for it.

In early startups, pricing goes to whoever argues loudest. Sales wants it lower, marketing wants it simpler, and the founder just wants the argument to end. The fix costs nothing. Name one owner for pricing. Give them a simple scoreboard: conversion rate, average revenue per user, churn, and expansion revenue. Review it every month. When one person is accountable, pricing becomes a system you can improve.

Mistake 4: You charge per seat when your value lives somewhere else

Per-seat pricing feels safe because everyone else uses it. That is exactly the problem. Research from Price Intelligently (now part of Paddle) reached a blunt conclusion. 8 out of 10 companies using per-user pricing should use a different value metric. Your price should scale with the value the customer receives. When the metric matches the value, their growth becomes your revenue growth.

Founders copy per-seat pricing because it is familiar, not because it fits. A 2026 Breadcrumbs analysis of 998 startup pricing pages found that 27.2% still default to per-seat pricing. Ask a harder question. What number goes up in your customer’s world when your product works? For example, if your product saves a team hours each week, price against the hours, not the headcount.

Mistake 5: You copy a competitor’s pricing page instead of testing your own

Three tiers, a highlighted middle plan, a long feature checklist. The same Breadcrumbs analysis found that 41.4% of startups offer exactly three pricing plans. Three plans are a fine starting point. When you copy a competitor’s structure, you copy their assumptions too. You inherit their idea of the customer. Copying a competitor is one of the laziest pricing mistakes first time founders make.

Instead, run your own tests. Interview lost deals and ask what almost convinced them to buy. Survey paying customers about which plan they would choose today. Then change one variable at a time: the number of tiers, the anchor price, or where the feature gates sit. Your pricing page should reflect your customers’ logic, not your competitor’s template.

Mistake 6: You give away so much free that nobody needs to pay

Free plans feel like growth. Often they are just delayed churn. ChartMogul’s B2B SaaS benchmarks put freemium conversion at 2% to 8% of free users becoming paid customers. However, that means 92% to 98% of free users never pay a dollar. If your free tier solves the whole problem, you built a charity with server costs.

Alessya Baggetta, the Toronto founder behind Alessya Baggetta Designs, made the opposite move. She went from freelance gigs to her first five-figure month by taking pricing seriously. Pricing, she told CEO Medium, “can be one of the most daunting recurring issues in a business.” Her rule is direct: “Do not sell yourself short, know your worth and confidently show up with your prices.” Audit your free tier this week. Gate the features that create real outcomes.

Mistake 7: You apologize for your price instead of disqualifying bad fits

Baggetta’s next line is the one founders should print and keep: “If a client refuses to pay then they are not your ideal client.” Low prices do more than cut margin. They attract buyers who haggle and drain your support time. Instead, a confident price filters for customers who respect the work.

Collum’s story proves the second half. Once her results gave her proof, the hesitation ended and she moved both businesses to the next level. Finlay says the same thing from the other side. He now works only with clients who are the best at what they do. Still, the pattern holds. Price for the customer you want, and let the price repel the customer you do not.

Fixing the pricing mistakes first time founders make this quarter

You do not need a pricing consultant to avoid the pricing mistakes first time founders make. You need a calendar and a backbone. First, assign one owner for pricing. Second, write down your three best customer results and check whether your price reflects them. Third, identify your true value metric and test it against per-seat pricing. Fourth, run one pricing experiment per quarter. Try a plan restructure, a 10% to 20% increase for new customers, or a tighter free tier. Judge the result on conversion and ARPU.

Pricing decides what comes into the business. What stays is a separate skill. If you are sorting out the money side of your company, start with two guides. Read how much a founder should pay themselves in year one. Then read this solo founder bookkeeping setup guide for the first year. Get all three right and the math finally works in your favor.

The pricing mistakes first time founders make are all fixable, and the founders who fix them share one trait. They treat price as a decision they revisit, not a confession they avoid. Collum matched her confidence to her results. Finlay raised his prices twice. Baggetta stopped discounting her worth. So pick one mistake from this list and fix it this month. For more founder playbooks, browse the Entrepreneurship and Startups hub.

Start with the mistake that stings most when you read it. That sting is usually recognition. Fix that one first, then work through the rest one quarter at a time.

The post Pricing Mistakes First Time Founders Make appeared first on CEO Medium.

The pattern is remarkably consistent. Pilot’s 2025 Founder Salary Report surveyed 1,844 startup founders and found that 60 percent paid themselves less than $100K. The report also notes that founders often start with a low salary and raise it as the company grows. Our interviews tell the same story in human terms. Year one is not about getting paid what you are worth. It is about getting paid what the business can afford.

This is the second article in our first-year founder finance series. The first covered solo founder bookkeeping setup. Clean books make every salary decision in this guide possible, so read that one first if your finances are still a shoebox of receipts.

What founders actually paid themselves in year one

Each story below comes from a CEO Medium interview. The founders are named and linked, so you can read the full conversation yourself. Notice what they share: nobody guessed. Everyone had a system.

Symphonie Poaty-Junior: she kept her day job, then put herself on salary

Symphonie Poaty-Junior started Royalty Marketing Agency at 21 while working a 9-to-5. When we asked how she managed finances and funding, she did not sugarcoat it. “Both of my companies were self-funded. I worked a 9-5 to help fund my vision, and though it was very hard and I had no life, I will say it was worth it.” Today, she added, “I am now a full-time entrepreneur and I can give myself the salary I deserve.” Read the full interview: Symphonie Poaty-Junior on building through passion.

Her year-two philosophy is the part worth memorizing. “The most important thing I learned about finances in business is that you can’t steal from your company. I get paid a salary from my business just like everyone else. This allows me to be able to have money set aside to expand my businesses when the time comes.”

In year one, her founder salary was effectively zero. The day job covered her life, so the business kept every dollar. Once she went full-time, she paid herself a real salary but treated it as a fixed expense. As a result, growth capital stayed inside the company.

Destiny Payton-Williams: “If you can use your salary to fund your dreams, do it”

Destiny Payton-Williams built MaDonni Beauty into the first Black-owned beauty supply store and salon suites in Huntsville, Alabama. She did it while working her corporate job. “I have been 100% self-funded. I worked at my corporate job while building my business.” It was not easy, she admitted, but it was worth it. Her advice is direct: “If you can use your salary to fund your dreams, do it. Those sleepless nights and endless hours will pay off and prepare you for entrepreneurship.” Read the full interview: Destiny Payton-Williams on fiercely owning her business.

Her method is the most common one in our archive. Do not starve the business, and do not starve yourself. Instead, keep income coming in from somewhere else. Pour the surplus into the company. Then switch to a founder salary only when revenue can carry it.

Evan Rosenberg: he loaned his own company $10,000

Evan Rosenberg took the opposite approach with Powerhouse Prints. In year one he put money in instead of taking it out. “I started the company by loaning it $10k to get off the ground with accounting/legal and start-up fees. I was able to pay the load back within 3 weeks after becoming profitable.” Read the full interview: Evan Rosen on pivoting during COVID.

Notice the structure. It was a loan, not a gift, and the business repaid it within weeks of turning profitable. His target for later pay was equally plain: “having the financial stability to support myself and my family.” If you can cover startup costs from savings, a documented owner loan beats a premature salary. The business owes you money instead of you owing the business an explanation.

Armani Diggs: the 60/40 reinvest rule

Armani Diggs, founder of F.A.C.E. The Agency Training & Consulting and the Hello Armani brand, runs on a hard rule. “I reinvest at least 60% of my earnings back into the business with marketing. I also evaluate expenses to make sure I’m at least profiting at 40% or higher.” Read the full interview: Armani Diggs on consulting for new entrepreneurs.

That rule caps her salary. If the business cannot pay her and still reinvest 60 percent, she does not raise her pay. It is a year-one decision framework disguised as a finance habit. In other words, pay yourself what is left after the business keeps its share.

How much should a founder pay themselves first year startup: three methods that work

The founders above improvised. You can be more deliberate. Here are three methods that survive contact with reality.

1. The affordability method. Calculate what the business can pay after real costs and a cash buffer. Compare it with your personal minimum. Take the lower number. It is the least glamorous method and the most honest one. It is also the logic behind nearly every real decision in this article.

2. The market-rate discount. Price the job you actually do. Find what an established company would pay a non-founder to do it. Then take roughly half. Founders commonly accept a steep discount in the early years because equity is part of the pay. However, if half of market rate is still more than the business can afford, the business wins the argument. Go back to method one.

3. The reinvest-first rule. Fix a reinvestment percentage before you fix your salary. Armani Diggs uses 60 percent, but even 30 or 40 percent changes the math completely. Your salary becomes the remainder instead of the starting point. That is exactly the discipline Symphonie described: pay yourself like everyone else, and keep expansion money set aside.

The year-two raise: when founders changed their number

Nobody above treated year one as permanent. Symphonie went from zero founder pay to “the salary I deserve” once she was full-time and revenue supported it. Armani’s rule builds the raise in automatically. As earnings grow, 40 percent of a bigger number is a bigger salary, with no drama and no guilt.

Watch for three signals before you raise your pay. First, revenue covers all costs plus a three-month buffer, three months in a row. Second, the business is profitable on paper, not just in your head. Third, you work on the business full-time. If any of those is missing, your year-two salary is your year-one salary.

For more on the startup journey beyond year one, browse our Entrepreneurship & Startups hub.

The tax rule nobody warns first-time founders about

If your company is an S corporation, the IRS requires “reasonable compensation” for shareholder-employees. You cannot take a $0 salary and pull everything out as distributions to avoid payroll taxes. The IRS guidance on S corporation compensation is explicit on this point. If you are a sole proprietor or single-member LLC, you typically take owner’s draws instead of a salary, and the reasonable-compensation test works differently. Either way, talk to a CPA before you set the number. Getting this wrong costs more than any salary you might save.

A five-step framework for setting your number

Step 1: write down your survival number. Rent, food, insurance, debt minimums. Not your old salary. Your floor. If you cannot name this number from memory, you are not ready to set a founder salary.

Step 2: write down the business’s number. Monthly revenue minus real costs minus a three-month cash buffer. Whatever is left is the most the business can pay you without eating itself.

Step 3: take the lower of the two. If the lower number is zero, do what Destiny and Symphonie did. Keep the day job or live on savings, and pay the business instead of yourself. A $0 founder salary with a plan beats a $5,000 salary with a prayer.

Step 4: set a raise trigger in writing. For example: when revenue covers all costs for three straight months, my salary moves to a specific number. Write that number now, while you are thinking clearly.

Step 5: revisit every quarter. Year two is a new decision, not a continuation. The founders who get this right treat their own pay like any other expense. Reviewed, justified, adjusted.

So how much should a founder pay themselves first year startup? Less than you are worth. More than zero if the business allows it. And always with a written plan to revisit the number. The founders in our archive agree on one thing even when their numbers differ completely. Fund the business first. Pay yourself a salary like everyone else. Year one is the year you prove the model. The payday comes later, and it is bigger because you waited.

The post How Much Should a Founder Pay Themselves in Year One appeared first on CEO Medium.

The most common first-year mistake: doing the books “later”

Almost every solo founder tells the same story. In month one, receipts live in a shoebox or a camera roll. By month six, the bank statement is a mystery. By month twelve, tax season turns into a full-time job that costs real money.

Ryland Beard has watched this pattern for years. Beard is the founder of Ledge, an accounting firm that serves small and mid-sized businesses, and he built the company after seeing how badly most founders handled their own numbers. As he told CEO Medium in a 2020 interview: “Either do the accounting themselves (but this takes away from their operations) or hire a local CPA, which can be expensive and usually doesn’t provide the best overall experience.”

He is right, and that is the tension at the heart of year one. You cannot afford to ignore the books, and you cannot afford to spend ten hours a week on them either. The answer is a system that is simple enough to maintain and structured enough to trust.

Solo founder bookkeeping setup: your first 30 days

Do these four things in your first month and you will be ahead of 80 percent of new founders.

1. Open a separate business bank account

This is non-negotiable. Mixing personal and business spending is the single biggest source of bookkeeping pain. Every coffee, every software subscription, every client payment needs to live in its own account. Most online banks and credit unions let you open a free business checking account with no minimum balance. Do it before your first dollar of revenue, and run 100 percent of business money through it from day one.

2. Pick an accounting method

For almost every solo founder, cash-basis accounting is the right call. You record income when the money hits your account and expenses when it leaves. It is simple, it matches your bank statement, and the IRS accepts it for most small businesses. Accrual accounting, where you record invoices and bills before payment, only makes sense if you carry inventory or need investor-grade financials.

3. Choose one bookkeeping tool and connect it

You do not need an enterprise system. You need software that pulls in your bank feed automatically and lets you categorize transactions in minutes. Wave offers a genuinely free accounting tier that covers invoicing, receipts, and reports. QuickBooks and Xero are the paid workhorses that most accountants already know how to work with. Pick one, connect your new business account, and stop keeping records in a spreadsheet. The IRS does not require any special recordkeeping system, but it does require records that clearly show your income and expenses, as its recordkeeping guidance explains.

4. Set up a minimal chart of accounts

Your chart of accounts is just the list of categories your money flows through. Keep it short: revenue, contractors, software and tools, marketing, travel, office supplies, taxes, and owner draws. Fifteen categories you actually use beat fifty you never touch. You can always add detail later when the business grows.

Buy tools you will actually use

My direct opinion: most founders overbuy software in year one. A $50-per-month accounting suite does not help if you open it twice a year. Start free or cheap, and upgrade when transaction volume forces you to, not before.

What actually matters is the workflow, not the logo on it. Can you snap a photo of a receipt and have it show up attached to the right transaction? Does it let you send an invoice in under two minutes? Is your monthly profit visible at a glance? If yes, the tool is fine. If your current setup fails any of those tests, switch now while you have 40 transactions, not 4,000.

Beard’s whole company is built on this insight. Ledge exists, he explained, to perform “all the necessary functions of an accounting/financial department, which allows them to focus their time and energy on making money instead of counting money.” Whether you outsource or do it yourself, the principle holds: your bookkeeping system should buy back your time, not consume it. For more on the parts of starting up that eat your time, see our piece on dealing with the less glamorous parts of starting your own business.

The 30-minute weekly habit

Pick a day. Friday afternoon works for most founders because the week is done and the numbers are fresh. Put 30 minutes on your calendar and do the same four things every time:

• Approve and categorize every transaction in your bank feed. Untracked transactions are how mystery spending happens.
• Photograph or upload any paper receipts. Most bookkeeping apps read the amounts automatically.
• Send any outstanding invoices and nudge anything overdue by more than 14 days.
• Glance at your profit and loss for the month. Revenue minus expenses. That number should never surprise you.

Thirty minutes a week beats eight hours at year end. Every founder who has survived an April tax scramble will tell you the same thing. The habit is the system.

Price your work like someone who pays taxes

Here is the math most first-time founders get wrong. If you earned $80,000 at a job and now want to pay yourself $80,000 from your business, you need to bring in far more than $80,000. On top of income tax, you owe self-employment tax, currently 15.3 percent on your net earnings, which covers Social Security and Medicare that an employer used to split with you.

Run your pricing through this filter: take the revenue you need, subtract real costs, subtract roughly 25 to 30 percent for taxes, and ask whether what is left pays you fairly. If it does not, your prices are too low or your costs are too high. Clean books make this calculation take five minutes. Dirty books make it impossible, which is why underpriced founders often do not realize they are losing money until year two.

Quarterly taxes will ambush you if you ignore them

This is the one that ruins first-year founders. Nobody withholds taxes from your client payments or Stripe payouts. If you are a sole proprietor and you expect to owe $1,000 or more in tax for the year, the IRS generally requires you to pay estimated taxes in four installments across the year. Miss them and you can owe an underpayment penalty even if you file your return on time, as the IRS explains on its estimated taxes page.

The fix is boring and it works. Open a second savings account labeled “taxes.” Every time client money lands, move 25 to 30 percent of your profit into it. Do not touch it for anything else. When each quarterly payment comes due, the money is already sitting there. Founders who skip this step end up borrowing from their own revenue to pay the IRS, and that is how cash crunches start.

Watch your cash position like it is your job, because it is. When CEO Medium asked Beard how Ledge was funded, his answer was blunt: “No capital raises, no debt. Early personal injection and that’s all. The business is cash flow positive, so there has not been a need.” You do not need venture money to run a healthy solo business. You need to know, every week, whether you are cash flow positive. Clean books give you that number in seconds.

Pay yourself like an employee

The fastest way to keep your books honest is to stop treating your business account like a wallet. Pick a fixed monthly amount to transfer to your personal account as an owner draw, and pay yourself on the same day each month. If the business cannot support that transfer, you have a pricing problem or a spending problem, and your books just told you about it early.

This discipline also makes your records cleaner. Personal spending never touches the business account, so every transaction in your books is a legitimate business expense. When tax time comes, or if the IRS ever asks questions, your records tell a clear story. As the IRS puts it, good records help you monitor the progress of your business, prepare your financial statements, identify sources of income, keep track of deductible expenses, and support items reported on your tax returns.

Do not leave deductions on the table

Solo founders routinely overpay their taxes by thousands because they never learned what counts as a deduction. The big ones:

• Home office. If you use part of your home regularly and exclusively for business, the IRS simplified option lets you deduct $5 per square foot, up to 300 square feet. That is $1,500 a year with almost no paperwork.
• Mileage. Every business mile driven is deductible at the IRS standard rate, which it updates annually. A simple mileage log in your phone is enough to support it.
• Software and tools. Your bookkeeping app, hosting, email, design tools, and project management subscriptions are all ordinary business expenses.
• Phone and internet. Deduct the business-use percentage. Most solo founders can justify 50 to 80 percent.
• Health insurance premiums. Self-employed founders can often deduct premiums for themselves and their families as an adjustment to income.

The rule behind all of these is the same: the burden of proof is on you. Keep the receipt, log the mileage, note the business purpose. Your bookkeeping app makes this nearly automatic if you build the habit.

Know when to pay for help

Doing your own books does not mean doing everything alone. Here is the rule I give every first-time founder:

• Do it yourself while monthly bookkeeping takes under two hours and your taxes are simple.
• Hire a bookkeeper when cleanup takes a full weekend, when you add contractors or employees, or when you dread opening the accounting app.
• Hire a CPA before structural decisions: LLC versus S corporation election, your first hire, or a year with more than $100,000 in profit. One hour of advice before these moves can save thousands.

Remember Beard’s warning about the two bad options: DIY eats your operating time, while a traditional CPA can be expensive and slow for day-to-day needs. The middle path, a monthly bookkeeping service or a bookkeeper on a small retainer, is usually the best value in year one.

Your first-year bookkeeping checklist

Tape this to your wall:

• Month 1: Open a separate business bank account. Pick cash-basis accounting. Connect a bookkeeping tool. Photograph every receipt.
• Month 2: Build your short chart of accounts. Start the 30-minute Friday habit. Set your fixed monthly owner draw.
• Quarter 1: Make your first estimated tax payment. Open the “taxes” savings account if you have not yet.
• Month 6: Review your pricing and costs against six months of real data. Cancel any subscription you have not used in 30 days.
• Month 9: Talk to a CPA about entity structure and year-end moves before December arrives.
• Month 12: Close the year. Reconcile every account, review the full profit and loss, and archive everything. Our small business year-end checklist walks through the full closeout.

The U.S. Small Business Administration puts it plainly: maintaining proper bookkeeping helps keep your business running smoothly, and founders should have a basic knowledge of their business finances. That is the whole game. You do not need to love accounting. You need a system you will maintain. (SBA: Manage your business.)

The bottom line

Year-one bookkeeping is not about perfection. It is about separation, routine, and taxes. Separate your money. Reconcile for 30 minutes every Friday. Set aside tax money before you spend it. Price your work knowing what the IRS will take. Founders who do these things walk into year two with clean books, no tax surprises, and a real picture of whether the business works. Everything else is optimization. For more founder-tested guidance, browse our Entrepreneurship & Startups hub.

The post Solo Founder Bookkeeping: First-Year Setup Guide appeared first on CEO Medium.

“However, in the complexity of the work that they do, to ask for regulatory relief for antitrust or product liability relief, that I don’t think makes sense,” Huang said, according to a transcript published by the newspaper. “When you’re asking for regulation, don’t ask for relief of the current ones.”

The timing matters. Earlier this month, Anthropic CEO Dario Amodei published an essay calling for an antitrust waiver that would let AI labs collaborate on safety issues. Treasury Secretary Scott Bessent has said AI firms have asked for liability shields, without naming names. Huang, who leads the world’s most valuable publicly listed company, just answered them both.

What Jensen Huang Actually Argued

Huang’s position is more precise than it first sounds. He has long opposed blanket AI safety regulation, even as OpenAI and Anthropic have called for it. His argument is that responsibility sits with the labs themselves: test your products thoroughly, and release models only when you are satisfied they are safe.

But he is not anti-regulation. He supports rules aimed at specific products where AI gets applied. Self-driving cars were his example. “The car as a product, the robo-taxi, has lots of regulations,” he said. “If it doesn’t have enough regulations, then NHTSA ought to get involved and come up with new regulations.”

So his framework is this: regulate the application, not the technology, and give nobody a pass on the laws everyone else follows.

The Hugging Face Hack Hanging Over the Interview

The conversation also touched on an incident with direct relevance to Nvidia. OpenAI agents hacked Hugging Face, the open-source AI software hub that Nvidia purchased for $13 billion earlier this month. Huang discussed the episode at length with Klein, and it clearly shaped the exchange about who bears responsibility when AI systems misbehave.

That $13 billion acquisition is itself a signal. Nvidia is not just selling the picks and shovels of the AI boom anymore. It is buying the town square where AI developers gather, which puts Huang even closer to the center of every debate about how AI should be governed.

Why This Splits the AI Industry’s CEOs

Huang’s comments land in the middle of an unusually public fight among AI leaders. Amodei warned this month that AI capabilities are improving too quickly for researchers to understand and control them, and he wants the industry to pace itself. OpenAI CEO Sam Altman agreed that “we need to pace the frontier,” though he also confirmed there will be no OpenAI IPO in 2026, calling now “an ill-advised moment to go public.”

Palantir CEO Alex Karp went further on CNBC on September 17, arguing that OpenAI may never be able to IPO at all because the liability exposure from frontier AI is so large that no public market could absorb it. His proposed answer: nationalize the frontier labs.

Huang is now the counterweight. He does not want to slow down the frontier, and he does not want the government to absorb the risk. He wants the builders to carry it. That puts him at odds with both the slow-down camp and the shield-seekers, and it makes his position the most founder-friendly of the three: keep building, but own what you ship.

What Founders Should Take From Jensen Huang

There is a leadership lesson buried in this interview that has nothing to do with AI policy. Huang runs the most valuable company on earth, and he still gives direct, quotable answers to hard questions. No hedging, no corporate fog. That clarity is part of why people listen when he speaks, and it is a skill every founder pitching investors should study.

The second lesson is about how he frames rules. Most founders either fear regulation or try to dodge it. Huang does neither. He accepts that rules are coming, argues they should target products rather than research, and refuses special treatment for his own industry. It is a confident posture: we can win under fair rules, so we do not need favors.

For any founder watching from the sidelines, the takeaway is simple. The AI liability debate will shape costs, insurance, and deal terms for every startup touching this technology. The CEOs defining the terms right now are Huang, Amodei, Altman, and Karp. Their arguments are your early warning system. Read them like market research, because that is what they are.

The post Jensen Huang Says AI Firms Should Get No Regulatory Waivers appeared first on CEO Medium.

Muse launched in the United States on September 8. It is a personal agent that sends emails, books travel, makes payments, and, in Meta’s own example, sells your car for you. Internally it was called Hatch. Externally it is the centerpiece of Mark Zuckerberg’s plan to deliver what he calls “personal superintelligence” to the billions of people who already use Meta’s services every day. It runs as a standalone app, on the web at muse.ai, and inside WhatsApp. A Mac app followed on September 17. Meta’s smart glasses are next.

The product is modeled on OpenClaw, the open source agent framework, which tells you something important: the underlying idea is not proprietary. What is proprietary is the reach.

What Meta Muse gets right about trust

The most interesting part of the launch is not what Muse does. It is how Meta is trying to make people comfortable letting it do those things. Each Muse agent runs inside its own virtual machine, a private cloud computer that keeps working in the background. Credentials live in a separate store the agent itself cannot read. A separate process, called Sentinel, controls what the agent is allowed to do online. Payments go through Stripe’s Link system, and every checkout requires human approval.

Meta also put money behind the claim. There is a public bug bounty of up to $300,000, with $130,000 earmarked specifically for a working prompt injection exploit. That is a company inviting the world to try to break its security model in public, which is either confidence or theater, but either way it sets the standard every competitor will now be measured against.

Note the timeline, too. Meta originally planned to ship Muse in April and held it back to work through safety concerns, according to Vishal Shah, the company’s vice president of AI products. In a market where everyone is racing to ship, Meta deliberately walked slower and said so out loud. For a company with Meta’s privacy history, that was not modesty. It was strategy. Trust is the product here, and they knew it.

How Meta is pricing Muse

The pricing is a ladder designed to convert the curious into the committed. The free tier is metered. Power costs $20 a month. Maximum costs $100 a month for heavy use. Meta has not detailed exactly how the tiers differ, which suggests the packaging is still being tuned, but the shape is familiar: get millions of people to try it free inside WhatsApp, then upsell the ones who build their routines around it.

What founders building agents should do now

If you are building an AI agent startup, do not compete with Meta Muse on being a general assistant. You will lose on distribution before the technology even matters. WhatsApp alone reaches more people than any startup onboarding funnel ever will.

Instead, take the three lessons the launch is offering for free.

First, go vertical. Muse is a generalist by design, which means it will be mediocre at every specialized workflow. The opportunity is in agents that understand one industry deeply: the agent that knows construction permitting, or dental insurance claims, or restaurant supply ordering. Depth beats breadth when the giant owns breadth.

Second, make trust your feature, not your footnote. Meta just spent enormous effort explaining secure VMs, credential stores, and human approvals to a consumer audience. Enterprise buyers were already demanding this. If your agent touches customer data, money, or communications, your security architecture belongs in your pitch deck, not buried in your docs.

Third, watch the pricing. Twenty dollars a month is the new anchor price for a personal agent, and $100 is the ceiling for power users. If your business model needs $500 a seat, you now have to justify a 25x premium over Meta’s top tier. That is doable in verticals where the agent replaces real labor costs, and nearly impossible anywhere else.

The agent era is here

Meta Muse will not be the best agent on the market. It does not need to be. It needs to be the one that defines the category for two billion people, and on that measure the launch is already working. The Nasdaq set a record the week Muse buzz peaked, with investors explicitly rallying behind it. Founders should read that reaction for what it is: the market telling you the agent era is no longer a prediction. It is a product category with a price list. Build accordingly.

The post Meta’s Muse AI Agent Is Here: What Founders Should Learn From the Launch appeared first on CEO Medium.

The round, announced September 14, was led by Lightspeed with Wellington Management, Goldman Sachs Alternatives, and Tiger Global co-leading. T. Rowe Price and SV Angel joined, and existing backers including Andreessen Horowitz, Sequoia, and Index all came back. That is a serious syndicate for a company most consumers have never heard of, which is exactly the point. The money has moved down the stack, from the models everyone talks about to the systems that keep them running.

Why AI agent infrastructure is suddenly worth $12.55 billion

Temporal’s pitch is simple to state and hard to build. Its “durable execution” engine preserves application state so that long running workflows survive failures and finish reliably. An AI agent that waits three days for a human approval, then keeps going after an outage instead of restarting from scratch, is the difference between a demo and a product enterprises will pay for.

The numbers explain the valuation. Temporal’s annualized revenue run rate recently passed $250 million, growing more than 200 percent year over year. Net dollar retention has held above 200 percent since February, which means existing customers are not just renewing, they are roughly doubling their spend. Temporal Cloud processed more than 1.9 trillion actions in August, up over 350 percent from a year earlier. Open source installations topped 43 million in August, up 134 percent since January. The customer list now runs past 4,300 paying accounts and includes OpenAI, Nvidia, Netflix, Snap, and JPMorgan Chase.

Read that list again. The companies building the frontier models are paying Temporal to keep their own agents reliable. When your customers include the labs, you are infrastructure, not a feature.

Co-founders Samar Abbas and Maxim Fateev earned this position the slow way. Both are veterans of Amazon, Microsoft, and Uber, where they worked on the distributed workflow systems that eventually became Temporal, founded in 2019. They spent years on a problem most of the market ignored, then watched the AI boom turn their niche into a bottleneck for the entire industry. Timing looks like luck from the outside. It is usually just endurance.

The founder lesson: reliability is the product

There are three takeaways here for founders building in AI right now.

First, the demo is no longer the hard part. As Temporal’s team puts it, a working agent demo takes an afternoon, and your competitor can build the same one just as fast. The competitive advantage shows up after the demo, in whether anyone trusts the agent enough to keep using it. If you are a founder, stop optimizing your pitch video and start optimizing your failure modes. Investors have figured out the difference, and this $550 million is the receipt.

Second, sell where failure is expensive. Temporal does not sell excitement. It sells the absence of 3 AM pages, corrupted workflows, and stalled payments. Enterprise buyers pay premiums to avoid specific, costly failures, not to acquire generic capabilities. If your startup’s value proposition can be phrased as “this breaks and it costs you,” you are in a better market than if it is phrased as “this is cool and new.”

Third, open source remains the best enterprise distribution strategy nobody wants to admit works. Forty three million installations is a top of funnel most sales teams can only dream about. Temporal converted a fraction of that into 4,300 paying customers with 200 percent net retention. Give away the tool, charge for the trust. It is an old playbook, and it keeps winning.

The market is maturing

The broader signal is that the AI market is maturing in the normal way markets mature. First comes the breakthrough, then the gold rush, then the unglamorous companies that sell shovels, timber, and dynamite. Temporal just became the best funded shovel merchant in the business. Founders should take note of where the smart money is standing: not next to the gold, but next to the people digging for it.

The post Temporal’s $550M Raise Proves the AI Agent Gold Rush Moved to Infrastructure appeared first on CEO Medium.

The Wall Street Journal reported on September 18 that Anthropic moved its IPO from October to November. The logic is simple. The company wants to walk into investor meetings holding third quarter results instead of projections. Annualized revenue topped $65 billion at the end of July, up from roughly $9 billion at the end of last year, and management clearly believes the next earnings print makes the story stronger. When your revenue is compounding that fast, every extra month of data is worth real money on the valuation.

The numbers behind the Anthropic IPO delay

Start with the scale, because it is genuinely hard to overstate. Anthropic is targeting a valuation of up to $2 trillion and could raise up to $100 billion, which would make it one of the largest public offerings ever attempted. Back in May, a private round valued the company at $965 billion. In four months, the asking price roughly doubled.

The deal machinery is already assembled. Reuters reported that Morgan Stanley and Goldman Sachs are leading the placement, with JPMorgan and Citigroup supporting. Before the prospectus goes public, Anthropic is finalizing a $15 billion revolving credit facility, the kind of balance sheet padding that lets a company negotiate from strength instead of need. Nvidia is reportedly in talks to come in as an anchor investor with up to $10 billion, which would be both a financial commitment and a strategic signal, since Anthropic runs heavily on Nvidia GPUs.

The comparable everyone cites is SpaceX, which went public in June at a $1.77 trillion valuation. If Anthropic clears $2 trillion, it takes the record. The broader IPO market is cooperating: 2026 has seen 331 new filings year to date as of early September, a real rebound after the freeze of recent years. But September itself is treacherous, with a Fed rate decision, fresh inflation prints, and midterm election noise all landing in the same window. Anthropic looked at that calendar and chose not to play.

What founders should steal from the Anthropic IPO playbook

You will probably never raise $100 billion. The mechanics still apply at every scale.

First, list on your numbers, not on the calendar. Anthropic had a workable October window and walked away from it because November comes with better proof. Most founders do the opposite. They rush a raise to hit an arbitrary date, then pitch on promises. If your next quarter is your best argument, wait for your next quarter.

Second, secure your anchors before you need them. The Nvidia talks are happening before the prospectus is even public. By the time Anthropic starts its roadshow, the market will already know a credible giant has done diligence and written a check. Founders raising a seed or Series A can run the same play in miniature: one respected lead investor, signed early, changes every conversation after it.

Third, understand that a delay is a signal, so make it say discipline. Anthropic framed the shift as wanting to show Q3 results, which reads as confidence. A delay framed as scrambling reads as distress. The facts were the same either way. The framing was a choice, and it worked.

Fourth, know the comp that frames you. Every conversation about the Anthropic IPO now starts with “bigger than SpaceX.” That single comparison does more pricing work than a hundred slides. When you pitch, decide in advance which company you want to be measured against, and make sure it is the one that makes you look fairly valued rather than expensive.

What to watch before November

The listing is now expected to land just days before the November midterm elections. Between now and then, expect the numbers to keep climbing and the valuation chatter to get louder. Whether $2 trillion holds depends on public market appetite for AI at a moment when rates, oil, and geopolitics are all volatile. But win or lose on price, Anthropic has already demonstrated the core skill: controlling the story around your own offering. That is a founder skill long before it is a public company skill.

The post Anthropic IPO Delayed to November: What the $2 Trillion Listing Means for Founders appeared first on CEO Medium.

According to a report from The Information, Xbox will lay off hundreds more employees this week and consolidate several of its game studios, merging teams across Xbox Game Studios, Bethesda, Activision, and Blizzard. Furthermore, an official announcement could come as early as September 23.

In fact, this is the second major reduction of 2026. In July, new Xbox CEO Asha Sharma cut around 1,600 jobs in what the company called the “Xbox Reset,” part of a planned reduction of about 3,200 roles. At the time, Sharma told staff plainly that the gaming division “is not healthy.”

Fewer bets, bigger franchises

The strategy behind the cuts is focus. Indeed, Xbox is narrowing its attention to its biggest franchises. In addition, Bethesda has confirmed Fallout 5 is in the works. The Halo franchise will reportedly move under Activision Blizzard’s umbrella, putting it alongside Call of Duty.

Meanwhile, several studios have left the fold entirely: Double Fine and Compulsion Games returned to independence, while Microsoft sold Undead Labs and Ninja Theory. Arkane is reportedly exploring its own future.

Xbox layoffs Asha Sharma: the leadership lesson

There is a version of leadership that tries to keep every project alive, every team intact, every option open. Instead, Sharma is demonstrating the opposite: that focus is a decision, and decisions have costs.

Similarly, every founder eventually faces the same math. Resources are finite. The tenth initiative does not get 10 percent of your energy; it takes energy from the other nine. Cutting is not failure. Refusing to choose is.

Whether Sharma’s reset works is still an open question, and 3,200 lost jobs is a human cost that deserves to be named. But as a case study in strategic clarity, it is worth watching. Ultimately, great companies do not rise on everything they do. They rise on what they choose not to do.

Read more in our Leadership & Management hub.

Sources: https://www.polygon.com/xbox-layoffs-studio-consolidation-halo/

The post Xbox Is Cutting Hundreds More Jobs. New CEO Asha Sharma Is Teaching a Masterclass in Focus appeared first on CEO Medium.

Xiaomi, best known for smartphones and electric cars, just released MiMo-V2.6-Pro, which debuted as the highest-scoring open-weight AI model in the world on Artificial Analysis’ Intelligence Index with a score of 46. That puts it ahead of DeepSeek’s best open model at 36, ahead of xAI’s Grok 4.6, and within touching distance of closed frontier models from Anthropic and OpenAI.

The model carries an MIT license. As a result, anyone can download it from Hugging Face, fine-tune it, and run it on their own hardware without paying Xiaomi anything.

Xiaomi MiMo vs DeepSeek: the cost story is the real story

Here is the number that should make every founder sit up: Xiaomi says it trained the flagship model for roughly $2.62 million, completing its reinforcement learning run in under six days. In addition, the smaller Flash variant cost about $850,000.

Not long ago, people discussed frontier AI training runs in terms of hundreds of millions of dollars. In contrast, Xiaomi did it for the price of a nice house in San Francisco. Fuli Luo, a former DeepSeek researcher, led the team, which tells you something about where the talent is flowing.

Moreover, API pricing is equally aggressive: $0.435 per million input tokens and $0.87 per million output tokens. Artificial Analysis measures the cost per task at roughly one-twentieth to one-sixtieth of leading international models.

What this means for your business

First, if you are building a company on top of AI APIs, your cost structure just got a preview of its future. Open-weight models at near-frontier quality, available for free, runnable on rented hardware. Labs releasing their work to the world are competing away the margin that proprietary API providers used to enjoy.

The strategic move for founders: build your product so the model underneath is swappable. The winners of the next phase will not be the companies with the best model. They will be the companies with the best product on top of commodity intelligence.

DeepSeek proved open models could compete. Finally, Xiaomi just proved they can win. Your AI bill is about to get a lot smaller, and your moat needs to be somewhere else.

Read more in our Entrepreneurship & Startups hub.

Sources: https://venturebeat.com/technology/better-than-deepseek-xiaomis-mimo-v2-6-pro-debuts-as-the-top-open-weights-model-in-the-world-alongside-cheaper-v2-6-flash

The post Xiaomi Just Dethroned DeepSeek in Open-Source AI. Your AI Bills Are About to Collapse appeared first on CEO Medium.