Don't Grind Into the Permanent Underclass
Recently I spoke to someone that asked me why I’m not interested in doing a 9-9-6 job at a startup in San Francisco.
It seemed obvious to me - why would I want to give away the best years of my life, my health, and my relationships away to a company that doesn’t guarantee significant upside to me?
But still, I hesitated to answer because there are other people who say that you should work hard in your 20s because you can take risks and recover more easily.
I also thought about my parents, who worked incredibly hard for many years of their life and spent hundreds of thousands of $ to send me to the US to, hopefully, have a good future.
And then to say “I don’t wanna make money because I don’t want to do 9-9-6” seems embarrassing.
Money or happiness?
It’s an undeniable fact that 9-9-6 equals burnout, deterioration of physical and mental health in exchange for money and clout.
As a first instinct, I believe the only time you should do it is if it was for your own company where you owned majority stake, because otherwise you’re not exposed to enough financial upside - adjusted for how risky startups are (90% failure rate) - nor is it guaranteed to be a mission you care about, which leads to moral & mental deterioration.
Many people say that even after they made money they are not happy.
It’s a story that’s told over and over. Yet, people still fall into the “rat race” of chasing the money.
“I think everybody should get rich and famous and do everything they ever dreamed of so they can see that it’s not the answer.” —Jim Carrey
Don’t get me wrong. I’m in just as much need of money as you.
”The permanent underclass”
However, is it a different story now due to AI?
Due to AI, I hear people say they must “escape the permanent underclass” as they believe that the edge to make money is going to disappear once AI can do everything.
Once AI agents can write all the code, build all the products, and sell the product, what are humans paid to do?
And if humans are not paid to do anything, where do they get money to spend?
The answer is
(1) other businesses.
(2) people who became financially free before AI does everything.
The rest of the population get Universal Basic Income (UBI) or some equivalent.
Two classes of products are sold: UBI products (canned soup) and luxury products (luxury soup).
The fear is that the wealth divide is widening and the time to make money is NOW before the AI incumbents settle in and the ladder to get to the upper class is kicked away.
Should you be scared?
“Permanent underclass” has good reasoning behind it. Over the past several decades, labor’s share of national income in developed economies has been in steady decline..
As computing became much cheaper and more capable, companies shifted their production methods away from labor and towards capital.
Combine this with offshoring, and it’s very clear why wages don’t grow as fast as company profits do.
Now, as we move towards zero-marginal-cost production, the same will happen. Wages go down, and some jobs will go away and never come back.
When this has happened in the past, as some jobs disappear, new jobs become available. For instance, software engineers during dot-com boom.
The problem now though, is that new jobs are not being created as fast - and the “permanent underclass” fear is that they never will.
The divide can be shown by the “K-shaped economy” diagram.
Businesses in the future will increasingly create things for elites who actually have extra money to spend.
Since AI does junior work, we only need senior architects who already have domain knowledge and judgement. Those that don’t already have access to tier-1 companies or universities are locked out.
Not THAT bad, but move quick
In reality though, it may not get as bad as “canned soup vs. luxury soup.”
This is why:
Firstly, the baseline quality of life goes up. If everything’s so cheap, everyone’s living well (the canned soup is actually really good).
Secondly,
We still need in person events, concerts, nurses, bartenders. A whole new class of jobs that I won’t try to predict pop up.
Even if you do nothing, achieving top-tier health, longevity, and deep relationships will be cheaper and more accessible than ever before.
What you need to do is
(1) survive the transition period as legacy industries collapse, non-AI asset prices collapse, and traditional white-collar wages collapse
(2) improve your position due to the various opportunities presented by AI-age uncertainty.
Gain as much agency as you can, while you can.
That difference causes things like prime real-estate, top-tier surgeons, educators, and retreats to be bid even further up by those that have AI agency v.s. those that don’t.
What should you do?
This brings us back to the question - is 9-9-6 warranted due to the risk of becoming part of the permanent underclass?
The answer, as it is for many questions, is a barbell strategy, which I stole from Nassim Taleb’s book Antifragile.
In investments, a barbell strategy allows you to protect your baseline capital from catastrophic losses while exposing a small portion of your portfolio to unlimited upside.
The loss is capped at 1x investment but gains could be 10-100x.
You avoid middle-risk which has capped returns and unlimited downside.
A barbell strategy has 3 components.
- Fragile component - breaks under stress.
- Robust component - survives stress.
- “Antifragile” component - improves under stress.
Examples of barbell strategies across life include,
- Exercise: 80% Zone 2 exercise that builds metabolic efficiency combined with 20% Zone 5 exercise that stimulates hypertrophy (overcompensation) and neurological adaptations. Avoid moderately exhausting workouts (junk volume) performed every day that leave you chronically sore and fatigued, yet fail to provide stimulus for peak strength and recovery.
- Time management: 80% highly-structured deep-work. 20% unstructured time for reading, trying new tools, networking, brainstorming with no agenda. Avoid semi-focused multitasking, constant meetings, messages and emails.
- Social life: 85% tight knit reliable core circle of family and friends where trust is guaranteed. 15% new events, or joining unfamiliar groups from different industries and backgrounds. Avoid the middle of transactional relationships or constant “coffee chats.”
For our purposes here, your strategy to avoid being stuck in the permanent underclass could look like this:
Antifragile component
Extreme exploration. Enter unfamiliar social networks and test new ideas. Tinker more with side ventures.
Dedicate a small percentage of your time, energy, capital to exploration. Some ideas:
- Launch exclusive, device-free, in-person dining or networking events in your city. As Linkedin becomes flooded with spam and AI-articles, the premium for physical events skyrockets.
- Use weekends to make and sell tangible physical items - carpentry, specialty foods, art, or sell your expertise as advisory services.
- Position yourself as an advisor that gets called when AI screws up. Benefit from screw-ups.
Take advantage of the Lindy effect. While machines become hyper-optimized, musicians that perform live sets, interactive seminars, actors on stage, become more valuable.
We pay extra for that, both with our money and our attention.
The Lindy Effect states that things that have already existed for 3,000 years are likely to survive another 3,000.
Examples: cultural traditions, music, storytelling, debate, crafts, human relationships. These improve and become more important with time.
If something has existed for 3,000 years of wars, technology, and fallen empires, it has proved that it is antifragile.
Instead of thinking about what’s going to change, think about what’s going to stay the same?
You can also deploy capital in your own projects or in small aggressive angel investments. Anything where potential loss is strictly limited.
Failure should be non-fatal - you can pick yourself up and try again, but potential upside should be limitless.
Another strategy is to use your capital to buy existing businesses that have cash flow and product-market fit. This addresses two problems:
- Commoditized products due to AI = more distribution competition. The solution is to buy existing distribution, or service existing ecosystems.
- Lack of defensibility: You own existing customers, proprietary data, human networks.
- Limited personal time & energy. You might burn out if you put only time into your projects. Solution: Put your capital as well.
As much as you can, keep 100% equity. You get the asymmetric exposure of a startup founder without giving up equity to VC or those that demand 9-9-6 hours.
Even if you fail, you rely on your income. If it works and you can scale, you work on it full-time and replace your income.
Robust component
Lower your baseline needs and have 6-12 months of savings. Zero debt. Turn a layoff into a minor inconvenience.
Remove more things - spending, information noise. While AI adds more (content, products, noise), you strip away. Value is in subtraction.
Make your income come from physical trades (electrical, plumbing, construction), hands on healthcare (nursing, PT), or in regulatory compliance. Any field where humans are needed.
Either that, or make your income come from a 40 hour/week job at a mid-size post-product-market fit company started in the last 5-10 years where you get to leverage AI.
Build relationships. Strong personal networks might be one of the only things that survive commoditization.
Cap work hours at 40-50 hours a week. Protect your time, health, sleep, relationships.
Invest your savings where you cannot lose capital during a market crash. Examples:
- Short-term treasury bills
- FDIC-insured cash reserves
- Debt-free tangible assets
Fragile component
Don’t depend on a stable paycheck from a mid-level digital desk job, or on a single employer. Don’t be like a turkey fed for 1000 days, confident and happy, until it’s eaten at Thanksgiving.
Avoid 80 hours a week job for slightly above-average salary because it consumes all your time without ironclad job security or significant equity upside.
Don’t go with trades that you believe are “certain” - e.g. the S&P 500. Yes, don’t invest in index funds. Why?
- You are forced to pay a premium to get into a crowded trade alongside everyone else.
- Every day of steady gains increases the confidence that the trend is permanent until an unexpected shock (Thanksgiving) causes a collapse.
Example: in the past, incumbents like IBM in the 1980s, Cisco during the dot-com boom, or Nokia in the mid-2000s seemed invincible right before a shift that made them worth much less.
Sure, you might say the S&P 500 is positive in the long run.
It only works if you have a true robust floor. Then you can invest without being forced to sell in a market crash.
Neither is S&P 500 going to give you outsized gains. It’s neither a safe bet, nor a risky bet.
Why barbell strategy is optimal
Now we can finally answer the question: is 9-9-6 warranted due to the risk of becoming part of the permanent underclass?
Nope, not necessary. Through the barbell strategy, we can live much happier, fulfilled lives that actually IMPROVE due to AI uncertainty.
I have huge respect for the sacrifice that startup founders and engineers make, and yet I don’t want their lives.
They act as an altruistic hero for society, but a sucker for themselves.
While society captures the upside of their experiments (not just from successes - we get information from failure too), most founders get lost youth, ruined relationships, chronic stress, and bankruptcy.
Don’t blind yourself with survivorship bias. The media celebrates the 0.1% of founders who bet everything and became billionaires (and attribute it to grit), while ignoring the thousands of founders that worked just as hard, sacrificed just as much, and ended up in ruin due to randomness.
The solution isn’t not to launch ambitious projects, it is to protect yourself from total ruin with the barbell strategy.
But Armaan, it’s too slow to do that, it’s not focused enough, and it doesn’t provide me with the desperation needed to succeed.
My rebuttal: focus and speed is useless when nothing is certain. You need maneuverability and the ability to pivot. Furthermore, desperation becomes negative (bad decisions) when you don’t have personal robustness (e.g. you can’t afford rent).
But Armaan, I don’t want to regret anything, or miss out on the AI boom.
My rebuttal: fair enough, but competing against the same metric as everyone else (hours worked, lines of code written, feature shipping speed) plays a losing game - like picking the overall worse option (everyone loses) in the Prisoner’s Dilemma.
Instead, focus on the Lindy effect. Play a completely different game. It’s more personally satisfying, and gives you an asymmetric advantage that still takes advantage of the AI boom.
You don’t miss out on anything, because you still dedicate 10-20% of your time to ventures.
If you currently work 9-9-6 at a startup, ask yourself - is this really the mission you want to give up the best years of your life for?
Is your life as a whole going to benefit from AI or are you becoming an uncelebrated martyr?
If you’re doing it to avoid the “permanent underclass,” is that the optimal strategy?
As for the question of whether I honor the sacrifices that my parents have made for me - I believe that if I become an underpaid, sleep-deprived clone in someone else’s venture-backed experiment, it would be the same as spitting in their face.
So the next time someone asks you to only work for their company beyond 50 hours a week - and leave you no time to do things on the side - because you should believe in the mission - remember that there are other options out there for you.
-Armaan @armaanagrawal_
If this resonated
Get my latest writing and thoughts delivered to your inbox.