Does Our Cognitive System Meet the Necessary and Sufficient Conditions for Success?

(Pure handwritten in Chinese, translated by AI)

When I first started out, I picked a few directions, including content creation, AI tech products, overseas outsourcing, and a 2B vertical small app on some overseas platform. I thought products and content were the most important ones. One brings traffic, the other brings MRR. The ones I looked down on the most were toB and outsourcing. I just genuinely felt they were low. But it turned out the only ones that actually touched money were outsourcing and the vertical small app. Not much, but real money.

Whether it's Golemancy, NanoWhisper, zza ai, or sidegpt, simply put, these are all features I imagined in my own head. This is the common disease of programmers, or the tech crowd in general. We take whatever tools we have in our hands, combine them into some feature, then try to push it in front of "the masses", fantasizing that someone will use it, someone will pay.

I was using a cognitive system stuck in the technical dimension, trying to find solutions in the multi-dimensional space that is business. Of course there's no solution. I'll get to the cognitive system thing later. First let me break down the mistakes I made.

The most fundamental problem with this "someone will pay" fantasy is that the "masses" we imagine in our heads are not actual humans.

Put another way, "the masses" is an abstract word. It's not a person. It's not a specific human being. This is exactly the set of concepts tech people don't understand: pain points, target audience, user stories, and so on. If a tech person's technical ability is the sufficient condition for building a product, then the product they build is only the necessary, not sufficient condition for becoming a business and making a transaction happen. Why is it not sufficient? Because there's still a huge gap between the product and the transaction.

PMF and demand validation are the prerequisites for verifying whether your business actually works. And many tech people don't get this. And that's not even enough. Tech people also don't understand transactions deeply enough. We need to give users a reason to pull out their wallet. This is psychology. We need to turn a one-time payment into a recurring payment. This is business model. We need to make users believe this problem is worth solving right now. This is behavioral economics.

AI will tell you to look at where users are complaining. But complaining is a statement, not an action. Paying money is the action. The piece tech people are massively missing is everything about money in a business. Those low-education bosses, those entrepreneurs full of MBA knowledge, they understand money very deeply and very practically. Behind some extremely simple questions, there are countless accumulated experiences and best practice cases. Pure technical thinking obviously isn't enough to make a business happen. This is my experience over these past few months.

Back to the cognitive system.

I work 12 to 18 hours every day. Luckily I didn't completely let go of the toB side, so I'm not starving. Ideals are ideals, reality is reality. It's not that the detours I walked through were unforeseeable. It's that from day one, I already knew I couldn't succeed in one shot. Every pit, I have to step into myself before I'll really believe it. What others say, in the end, you don't really get it, you don't really feel it deeply. And depth itself is a kind of wealth. I believe in the end, the logic of getting things done is a contest of who understands certain things more deeply. So now I don't post short content. I firmly believe deep content is the kind that has a long tail effect, the kind that keeps producing value.

Can I, before doing something, judge what percentage of this thing I actually understand? In other words, is my cognitive system enough to support this thing? Say I can only see 60% of this thing. Then what is the remaining 40%? And the deeper question is, how do I even know that what I see is 60% of it? I don't even know the evaluation standard. I don't even know which angle to look at the thing from. So how do I evaluate? Who do I find to evaluate? What do I evaluate?

This series of questions is itself the problem, not the questions above. It's a bit twisty, but let me try to explain it clearly.

We need a methodology for finding methodologies.

There is nothing new under the sun. I've wondered if our universe is just a molecule inside some giant world. I've wondered what the world would look like if you removed the monetary system. I've wondered if marriage and law are inventions, and then whether language is also a human invention too. And none of this came from any book I read. It really came from reasoning freely in my own head. But all of these things have already been thought through by certain people, in some time and place, long ago.

People often say "don't reinvent the wheel". And yet some smart people always make this exact mistake, going off to rethink solutions from scratch.

If we don't make use of other people's wisdom, that itself is a kind of arrogance.

So the methodology for finding methodologies comes from a definite profession, a definite business model, a definite book. These definite things, in terms of probability, raise the probability of your thing succeeding. They add a little bit of certainty in the uncertain business world. This is math. This is science. For anything you want to make happen, you should go find the benchmark. If you want to think everything up from scratch on your own, you'll be exhausted to death. You trust your own wisdom too much. This is arrogance, let me say it again. This is the common disease of smart people. It's a mistake.

Look at who in the world is doing similar things. If they exist, what's their business model? What are their marketing methods? How do they abstract who their product is for? How is their marketing funnel designed?

What roles are usually needed to do this thing? What skills do these roles, or these professions themselves, require? What methodologies do they need to master? At this point you don't need to think too deeply about why the methodology is designed this way. It might be hundreds of years of accumulated best practices. It might be blood and tears lessons from generations before us. But it's already been proven to work. No matter how smart you are, you should first use these mature methodologies other people have built up.

The only thing we need to do is find it. Find those methodologies.

Let's call our methodology for finding methodologies the meta-methodology. Then from a philosophical angle, we need to ask a few fundamental questions, to determine what I said above. Is my cognition sufficient? What percentage of getting-things-done does my cognitive system actually satisfy?

You can have the following discussion with AI.

The Thing itself:

• [Thing] In the existing world knowledge system, which field does it belong to? What abstract concepts does it have?
• [Thing] What is the structure of the thing itself? Is it part of some bigger structure? Where does it sit inside that structure?

People:

• [Thing] What roles need to be brought in? Which are the key roles? Which are the secondary roles?
• [People] What abilities do they need to have? What is the methodology behind these abilities?
• [People] How does information flow between these roles? Where does each role sit in the whole chain?

Benchmarks:

• What's the current pattern of this thing?
• What are the existing solutions? (people, things, processes, structures, all are fine)
• What are the existing best practices? (people, things, processes, structures, all are fine)

The above is some of my thinking on the cognitive system level over these past few months. Doing one thing well isn't the actual problem. But if our cognitive system itself isn't enough to support our business, we have to upgrade it.

Cai Yongji here, wishing your future better and better. Hope this article helps you.

Likes and comments are welcome. If you have different opinions, criticism and corrections are welcome.