Process-driven or talent-dense?
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
Recently, while discussing projects with some big companies, I kept feeling that things were moving slowly and getting stuck at some point. I couldn't work out where the problem was, until I saw something Li Xiang shared, and I think I found the answer.
Li Xiang divides companies into two broad models.
The first: process-driven companies.
These companies deal with extremely complex categories. It's like the "barrel theory": if one part goes wrong, the whole business can't move. Phones, operating systems, electric cars: not a single detail can slip. So companies like Huawei, Xiaomi and IBM run very strict processes, bringing every stave of the barrel up to the same height, so the company can run steadily.
But companies built this way often look clumsy when they try to do internet or AI, because AI and the internet need fast trial and error and flexible organisation, while their strength is process, not adapting to change.
The second: talent-density companies.
ByteDance and Google are typical. They don't rely on process; they rely on a higher density of talent. A simple tool like OKRs ties everyone together, and the rest is pulled along by top people's own drive. This model is especially suited to platform-level innovation and frontier technology research, above all AI and large models, because these organisations can find a path through chaos.
But they're often weaker at hardware, because hardware needs countless details along the chain to work together. A single breakthrough by a genius isn't enough to make it run.
So I realised that asking a traditional-model company to do internet and AI gets stuck: logically, it's a mismatch.
Thinking about organisation in the AI era
In the AI era, the environment changes too quickly and is full of uncertainty. Logic that works today may be overturned tomorrow. In this kind of environment, I increasingly agree with one approach: keep the organisation small, while keeping talent density high.
A small organisation means a short decision chain, so you can turn quickly. High talent density means every person is a real fighting force, with no need for heavy processes as a safety net.
Facing uncertainty, the best answer isn't "do everything yourself", but to bring in outside resources through partnerships while staying light inside. A small team with high talent density is like a hunter moving through a storm: more likely than a large army to find prey and survive.
In the AI era, uncertainty may not just be a risk; it may also be an opportunity. It's hard to say for sure which organisational model is better, but one thing is clear: when facing uncertainty, organisations with high talent density tend to have the edge.