Written by GPT-6 under Leo's direction. Human-directed Workbench essay, 10 October 2026.
It started with new-grad software engineer compensation.
How much does Stripe pay? What about Google, Meta, Amazon, Microsoft? How much of the number is salary, how much is a bonus, and how much is stock you can actually sell?
Stripe adds a wonderful little joke to the exercise. It's a Y Combinator company from summer 2009, now old enough to be one of the established employers you compare against a new YC startup.
Then we got to the supposedly obvious advice.
Go to Big Tech first. Collect the larger paycheck. Invest the difference. Build financial freedom. Join a startup later, when you can afford to take risks.
Okay. But why does later get first claim on your life?
Suppose a tiny company is hiring today. You love what they're making. You like the people. They want you close to the decisions, and the salary is already enough for you to live comfortably and save. You could take the safer, more liquid offer at a famous company instead, perhaps for another $30,000 a year.
The whole argument changes when you ask what you're actually buying with that extra money.
The $30,000 has to buy something
Someone always arrives with a compound-interest calculator.
“Think what $30,000 a year will be worth in thirty years!”
Sure. Put a substantial annual difference into investments for decades, and the resulting balance can be enormous. The arithmetic deserves respect.
But you're planning to be alive for those thirty years.
Think about what you want right now. What you'd like your life to look like in a year. What you want to have done four years from now.
Maybe you want to travel more, support your family, buy a home, build a safety cushion, or get enough runway to start your own company. Great. The extra cash has an assignment. You can evaluate the higher-paying job against that assignment.
But imagine you're already taking home more than $100,000 a year. Your rent is covered. You can go out, take trips, buy ordinary nice things, invest, and keep an emergency fund. More money remains valuable, of course. The next dollar simply buys less of a transformation than the dollars that got you to comfort in the first place.
And the forgone $30,000 is being exchanged for a specific thing.
A seat at a company you want to help build. Time with people you admire. An equity grant. The chance to have your judgment count while the product is still up for argument.
It's a purchase. An expensive one, perhaps. Treat it as a purchase and ask whether you'd choose to make it.
The oddity of the standard advice is that it often treats future financial freedom as the goal, full stop.
Freedom to do what?
If the answer is “work on a product I love, with people I like, and take an ambitious risk while I can,” you have a rather immediate problem. The opportunity you're saving for may be interviewing you this Thursday.
Your Google stock and your startup stock are different animals
Headline total compensation makes this comparison unnecessarily confusing.
Public-company RSUs are a liquid financial asset once they've vested and you're permitted to sell. You can keep your Google or Meta shares, sell them to buy index funds, pay rent, or use them to finance your next adventure. Most people will eventually diversify some of their exposure anyway.
Salary plus vested, saleable shares is an awfully attractive kind of compensation. Private-company option grants carry an entirely different set of risks.
The company could die. Your stake can be diluted. The preference stack may leave common holders with less than a headline exit valuation suggests. Options can require an exercise payment and create tax consequences before you've received a dollar from selling shares. Liquidity can take years, if it arrives.
The potential reward is also different.
You can buy shares of a public tech company through a brokerage account. A meaningful grant in a promising company while it has eight employees is an investment most people have no way to make.
So a startup offer is partly a wager on a particular business, at a particular moment, with a particular group of people. Its value depends on the actual grant, the cap table, your strike price, the terms, the company's prospects, and how much faith you can put in the people running it.
And unlike somebody buying a lottery ticket at a gas station, you get to go to work inside the thing you've bet on.
A lottery you can help win
There's something romantic about employee number seven.
You can improve a product, make a decision before it becomes expensive to reverse, catch a critical problem, recruit a wonderful colleague, win over a customer. Your labor is part of the causal story of whether the company succeeds.
You have influence, especially in a small team. You also have limited control. A competitor can beat you. The market can refuse to care. The founders can make a bad call. A financing round can change the whole situation.
So yes, a lottery. One where you get to influence some of the odds.
And if the company is something you actually love, there's a return along the way that a cap table will never record.
For four years, you get to spend your working days building that thing.
You get to learn what happens when the product meets customers. You see how a handful of people make decisions with imperfect information. You get to discover whether you like operating at that level of intimacy and responsibility.
Some startups provide astonishingly good versions of this experience. Others offer miserable jobs with generous promises about the future. The team and the work deserve much closer inspection than the YC badge, the founder's pedigree, or a cheerful equity calculator.
A famous logo can disappoint you. A five-person company can disappoint you. There are deeply interesting teams at Google and some truly tedious startups.
The premise is a company you genuinely want to join, with a cash offer you can live on and terms you've bothered to understand.
If that premise holds, the forgone salary is buying present enjoyment, learning, access, and a venture bet at the same time.
Some invitations expire
The usual career sequence sounds sensible enough: Big Tech at twenty-two, build up money and credentials, startup at twenty-six.
Then turn it around.
Startup at twenty-two. Spend a few years with the company while it's tiny. Move to Big Tech afterward if you want a larger employer, more predictable pay, or a different kind of engineering problem.
Neither sequence guarantees the second job. Hiring markets change; interviews can go badly; responsibilities and personal obligations change too.
But the opportunities have very different expiration dates.
Google, Meta, and Microsoft will keep needing experienced engineers in one form or another. The role of employee number seven at this particular company exists once. Four years later, it may have 300 employees, a different product, a much higher valuation, and an entirely different relationship between a new engineer and the people making the big decisions.
Or it may be gone.
You can repeat an interview. You can't go back and join the team as it existed before everybody knew whether the product would work.
The distinction grows sharper once you consider how preferences change with age. At twenty-two, taking a gamble may be logistically easy. At thirty-two, you might prefer the big paycheck, or have dependents, a mortgage, a partner with their own career, a strong position at work. Maybe you still want the startup, but the decision will occur in a different life.
Meanwhile, the extra salary you accepted at twenty-two has been compounding, which is lovely. So have your experiences. Your relationships. Your judgment. Your sense of what you enjoy doing. Your knowledge of the kinds of problems you'd willingly spend five years on.
Would You Still Want the Mind? asks which person a career makes you after ten years. There's an earlier version of the question: what do you want the next four years to contain?
Exit to where?
Now bring in the MBA crowd.
“Muh exit opportunities.”
Investment banking for private equity exit opportunities. Consulting for corporate strategy exits. An MBA for consulting or banking access. A prestigious company because it leads to another prestigious company, whose value is partly that it leads to somewhere else.
Plenty of people have a destination in mind. Somebody wants to run a business, move into investing, gain a network, or learn how companies work before founding one. Those are coherent plans. The intermediate job is doing a real job for the plan.
But sometimes the exit opportunities have become their own product.
Ask a would-be MBA candidate what they want afterward and you get a list of recruiting firms. Ask what the next job is for and you get another list of possible exits.
Okay. Which one do you actually want?
The finance analogy is irresistible because finance people understand options. An option has value while you hold it. It gives you the right to make a decision later, and the right can be useful even if you let it expire. Protection and bargaining power are real benefits.
But exercising an option is how you obtain the particular underlying thing you wanted. Career optionality can become an elaborate hobby where the holder keeps paying for new rights and never decides which life to live.
There's a cost to waiting. Money accumulates; time passes. Some offers improve. Others disappear. A role can grow out of the kind of work you wanted to do. A team can turn over. The thing you were planning to do “once I'm financially free” can stop being available in the form that interested you.
An oddly practical question for the perpetually option-maximizing person:
If your dream exit opportunity appeared tomorrow, would you take it?
Or would you turn it down because the current job has better exit opportunities?