Jingle bell 初号机

PRD is dead, the era of super individuals has arrived (Part 2)

This means that a good PM: has good product thinking → can create good products
A bad PM can also create product prototypes

But due to a lack of systematic thinking and product sense, a bad PM generates a bunch of useless features + terrible design using incorrect requirements

Coupled with the traditional psychology of "since we're here"
it's easy to casually launch a bloated product with declining quality
So having systematic thinking is very important

Because in the future, there will only be two types of people: Builder or Reviewer
Builder: responsible for AI + product + design + engineering
Reviewer: responsible for system architecture + product strategy

Sequoia Capital partner Julien Bek wrote an article, which essentially says: the next trillion-dollar company will be a software company disguised as a service company

The software industry is entering the "era of super individuals", and individuals are as well

But the real threshold has never been technology, but execution

Because the vast majority of people are not incapable, but are constantly avoiding

Just like if you are someone who really dislikes doing housework, and you currently have an important and difficult task that you need to implement, but because you instinctively fear executing this difficult task, your brain might even say: I haven't done housework in a long time, should I go do housework first?

See, as long as you are not forced to face what you least want to do, your brain might even choose to accept a second least favorite task, this is an avoidance mechanism

As long as you are not currently executing the goal that needs to be completed, you are avoiding

So some people seem very smart, and work hard every day
But all assumptions are completed in their minds, they aim too high, feel that everything is easy, when friends around them succeed, they think for a moment and say what's the big deal, I understand the framework, I can do this too, but after a few months, they still haven't done anything

Why is this the case?
Because execution is risky
Once you really do it, you might fail

So to avoid this failure, many people choose to remain in the preparation stage forever: as long as I don't start doing it, I will never fail

And they can maintain an image of: looking very impressive, speaking confidently, knowing everything, giving pointers on everything, and as long as they want to do it, they can succeed

Speed + execution

The PRD is dead, the era of super individuals has arrived (Part 1)

Before the emergence of Claude, the software development process was typically as follows:

Propose an idea → Product Manager writes PRD → Designer creates UI → Technicians write code based on UI

The reason for this is that writing code is very expensive (time + manpower)

Thus, the software industry has formed a professional division of labor:

Product: Understand requirements
Design: Create interfaces
Engineering: Write code

And the PRD is the core of communication among these roles.

However, with the development of the AI era, Vibe Coding can directly turn ideas into functional software.

This means: the PRD is dead

Because now: Idea → Prompt → AI → Product prototype

The cost of implementation has been greatly reduced.

In the past, the biggest bottleneck was: writing code

Now, the biggest bottleneck has become: Review

Which means: determining whether the code is correct, determining whether the product is reasonable, determining whether the system is usable

The PRD has not truly disappeared; although the traditional PRD is dead, requirement descriptions are still very important.

The future form of the PRD may become: Prompt, Prompt context, Prompt architecture.

In the future, everyone must become a “Generalist.”

This means that regardless of whether you are currently a PM, designer, code engineer, or founder,

You need to become a Generalist

With capabilities in: Product Sense, Engineering Sense, Design Sense

Because nowadays, communication costs have become the largest cost.

A Generalist + AI = the original three-person team (Product Manager + Development + Design)

Super individuals are emerging. And these people have always been valuable.

Recently, a senior student created a project in 10 days and received 30 million investment, and the news went viral. His experience and insights are documented here👇: "In 3 months, from the micro-influencer BettaFish open-source project exploding to receiving 30 million investment, I experienced the super individual era brought by Vibe Coding". Just search for it.

In the future, Coding Agent will become a fundamental skill.
And the real core capability will become: systems thinking.

Because now the execution costs have dropped significantly, you just need to tell the AI what kind of product you want to make, and it will quickly help you generate a prototype.

How Ordinary People Can Truly Make Money through AI

First, let go of anxiety. AI has not created new business models; it has merely maximized the efficiency of old models.

Next, I will elaborate on the fifth point:

Even if you deploy crayfish and set up Claude Code, it doesn't mean you can make money.
What truly makes people money is always the business model, not AI.

A friend of mine, Xiao Y, who does cross-border e-commerce on TikTok (has left ByteDance), recently showed me an analysis page of his own Vibe Coding on TikTok.

The purpose is to capture the total sales of single products in the European and American regions on the daily chart / 7-day chart.
This friend sourced products from 1688 and sold them through TikTok.

With this panel, he no longer needs to manually scrape data, study bestsellers, or analyze the market; AI has helped him improve efficiency.
He only needs a self-coded panel to see everything clearly.

Buy high and sell low, it’s simple; you immediately understand what Xiao Y is doing and how he is making money.
But knowing and executing to actually making it happen is separated by more than just eighteen thousand miles.

Moreover, can you say that AI is what made Xiao Y earn money? It’s not; the act of buying low and selling high for profits overseas has existed since ancient times on the Silk Road.

What has always made Xiao Y money is the cross-border e-commerce model. He has run through the entire process, from shipping channels to maritime processes, from obtaining qualifications on TikTok to successfully selling products and handling after-sales, every step is not easy.

AI has only improved our production efficiency, helping us save more time.
But what truly makes us money is all the old models.

Business models boil down to 3 things:
Why people, what problems to solve, and how to charge.

For thousands of years, it has been these 3 things, unchanged.
AI has brought about a leap in productivity, but business models have never undergone a qualitative change.

Many things seem new, like AI companionship, AI children's education, AI customer service.
But if you take a closer look, the model has never changed; whether it’s emotional needs or maternal and infant education, the demand has always been there, and the business model has never changed, only the unit service cost has been decreasing.

So, in the next 5 years, the speed at which AI creates millionaires will be much faster than the internet.
But the premise for realizing this profit remains:
You need to run through a complete business model.

These two pieces of news look very magical together

First: ByteDance's North America Seed team is offering recent graduates an annual salary of 1.8 million dollars, 10 million RMB

Second: Anthropic released the "AI's Impact on the Labor Market Research Report" on March 5

The core conclusion is: AI has not significantly increased unemployment 😂

The report introduced a new metric: AI Exposure

Simply put, it refers to the degree to which a profession is affected by AI

1⃣ Professions most easily impacted by AI, see the chart

Their common characteristic is a high reliance on text, logic, and information processing

In the past, automation primarily impacted blue-collar manufacturing

But AI's first wave of impact is on: highly educated knowledge workers

2⃣ Employment hasn't decreased, but it's harder for newcomers to enter the industry

The report also contains an important statistic: the entry rate for newcomers aged 22-25 has dropped by about 14%

In other words: companies are not undergoing large-scale layoffs, but they have started to stop hiring new employees

Reduced hiring → Organizational downsizing → Some positions disappearing

3⃣ The impact of AI on employment is not a sudden explosion, but a gradual compression of entry points

And as illustrated by the example of the multimillion-dollar salary at the beginning, the labor market in the AI era is showing a very extreme structure:

On one side: ordinary white-collar workers whose efficiency is improved or even replaced by AI; on the other side, the prices for top AI talent are exploding

Within five years, AI will create more millionaires than the internet did in 20 years

Statistics even show that in the past six months, AI has already birthed 190,000 new millionaires, averaging 1,000 per day

A few days ago, there was a more ironic discussion in ByteDance's internal group

4⃣ The programmer community is the most contradictory group in the AI era

For the past few decades: countless programmers have open-sourced their best codes, which make up the GitHub open-source framework; AI has grown up consuming these corpora

As a result, once AI learned: the first to be optimized out were programmers

Without open-source, the barrier to programming might be as high as that of doctors and lawyers

5⃣ Career growth paths are disappearing

The typical growth path in an industry used to be:

Newcomer → Junior → Mid-level → Senior

But in the AI era: Newcomer + AI ≈ Mid-level Engineer

Companies no longer need to train newcomers

Opportunities for ordinary people to enter good industries are decreasing

Sold 2 BTC at a price of 88000 on the delivery date at the end of November 28, premium 1480*2=2960

If BTC drops below 85000 by the end of the month, I will fulfill my wish to receive 5 large coins. If it does not drop below 88000, I will still receive over 5000 in premiums. Regardless of whether it rises or falls, I am happy.

The premium is like "insurance" in options trading.

You sold a put (a put option) = You promise "If the Bitcoin price drops below 88000 dollars, I will buy it for someone else at this price."

The buyer pays you a fee in advance for this "insurance" = the premium, which is 1480 dollars for each Bitcoin as mentioned in the post.

Regardless of whether it ultimately drops or not, this money is yours —
If it drops, you buy the coins (you believe in Bitcoin anyway, and if it drops below 88000, you will buy it).
If it doesn't drop, you just earn this money for free.

It's just like selling insurance; customers pay you a premium.
If nothing happens, you earn, and if something goes wrong, you lose money.

If you are bullish on Bitcoin, you can use this as a hedge.