During college, my class was instructed to study Schopenhauer in political philosophy.
Complete bummer, pessimist, the guy just makes you want to kill yourself.
There was one quote about him though that really opened my eyes.
Schopenhauer's philosophy is the mirror of his own nature... What he saw was not the world, but himself writ large." Nietzsche
Schopenhauer basically talked about himself. Not the world. His prior conclusions on how the world works influenced his work. However, you and I know this is not the correct scientific method to apply.
You have to start from how the reality works and then derive your work.
We fall victim to this everyday.
Humans are so good at interpreting new information to justify their prior conclusions.
Which brings me to AI replacing services.
There is a widespread belief in tech circles that AI is a bigger opportunity than the web itself.
Why?
Because it literally replaces humans and services industries.
Here is one chart from Greylock (no offense to them) depicting this hypothesis:
The idea is that if AI can replace services jobs performed by humans, you can replace the employee expenditure with tech expenditure in form of software revenue.
This brings about 100 to 200x expansion of market size for software and AI companies to capture.
You got excited, didn't you?
I disagree.
Not based on belief but based on experience.
There are two reasons why I disagree:
The narrative by VCs The customer's expectation
1. Narrative by VCs
Now we know that VCs were sort of in a crisis before ChatGPT's launch.
2021 had happened where VC FOMO and greediness had skyrocketed with many fraudulent companies getting funded not based on merit but "vibes".
When market crashed, this broke LP trust, spiralling into a fundraising crisis for venture capital firms who make more fees when they get more funding.
In short, the incentive is:
More LP capital raised → more fees
So VCs needed a new narrative.
ChatGPT became the saviour.
The narrative this time?
AI will replace services expenditure, meaning that the tech market is 100 to 200x bigger. This is bigger than the web itself!
You see, noone knows. It is not a done deal. However we act as if it is certain.
Why?
Because incentives. We are projecting our prior conclusions, which we favour, on how the world will turn out. Instead of understanding the world and only then arriving at conclusions.
VC incentive is to make every investor out there to believe that this is the biggest opportunity in history.
You are either in or out.
And you do not want to be the guy who missed AI.
So you give them money. They take fees.
Problem solved.
2. Customer's expectation
When I was running Godmode, I was one of the first champions of the idea of AI replacing services.
That is why we priced our software in credits.
Noone wanted it.
I would demo the product, then the customer would spend 10 minutes asking questions about pricing.
"How much could it cost?"
"What if I don't get a reply?"
"What if the output was bad?"
This created too many complications that they would leave the call, not even knowing the price.
Then we changed the pricing.
→ $250 per month for 250 leads. $1 per lead. Whatever else you want.
Customers stopped asking questions.
It was New York pizza cheap. Simple. They knew how much they would be spending at the end of the month.
Smooth sailing.
Then I started doing the math. Our value proposition is replacing a lot of mediocre sales people, isn't it?
Let me walk you through what this really means.
A salesperson costs $4,000 per month. That's $48,000 per year. Not including benefits, office space, equipment. Just base salary.
What do companies normally pay for software they love?
Slack: $12.50 per user per month
Salesforce: $75 per user per month
HubSpot: $800-$3,200 per month for a whole team
Even the most expensive enterprise software rarely exceeds $500 per user per month
Now watch what happens when we try to capture that $4,000 salary with AI:
Let's assume a salesperson manually writes unique emails by hand and really crafts them.
He or she can write (including research) 20 a day.
For 22 working days, that is 20 × 22 = 440 emails or outreach messages per month.
For Godmode's pricing at $1 per lead, that is about $440 revenue.
The salesperson doing the same work costs $4,000.
We capture $440 out of $4,000. That's 11%.
Eleven percent.
Not 100%. Not 200x. Eleven percent.
Why? Because customers don't see AI as a replacement for their people. They see it as software. And software has different rules:
It must be predictable in cost
It must be cheaper than humans
It must be reliable
It must fit into existing software budgets
The moment you price AI like a human replacement, customers run. The moment you price it like software, they buy. But software pricing means you capture a fraction of the human cost you're replacing.
Here's the truth no one wants to admit:
The AI-replaces-services narrative is fantasy.
Not because AI can't do the work. It can.
But because the economics don't translate the way VCs claim. When you replace a $50,000 employee with AI, you don't capture $50,000 in software revenue. You capture $5,000 if you're lucky.
The market isn't 100-200x bigger. It's maybe 10-20% of what the PowerPoints claim.
Need an example? Good. Coding.
You must be paying your software engineers around $100,000 yearly.
Now that vibecoding is out there, when was the last time you committed to pay $100,000 to Lovable or Replit or Claude?
This doesn't mean AI is worthless. It means we need to stop lying to ourselves about what it's worth. Build AI products that customers want to buy at software prices. Stop pretending you'll capture human salaries.
The VCs need you to believe their narrative to raise their funds and collect their fees.
You need to understand reality to build a business that survives.
Choose wisely. All this is not to say AI does not replace labour. It does.
Question is: does that matter “for you”?
Focus on abundance, not replacement.
MD out.