AGI lets machines think; AGP lets machines do

In recent years, the embodied intelligence industry, where AI large models are applied, can be described as advancing by leaps and bounds. From VLA to various world models, we have been trying to use AI's successful path to achieve machine intelligence in embodied intelligence. But Intenite believes that AI describes machines that can think, while the vast majority of human civilization's output has never come from "being able to think," but from being able to do: doing something stably, repeatably, and correctly in the real world.

Thus a gap has emerged: we are desperately making machines smarter, yet we rarely discuss where productivity comes from after intelligence. Furthermore, in real social production, can embodied intelligence be invoked as productivity by everyone, just like using today's AI?

Our answer is: General Productivity.

What is General Productivity?

Intenite believes: the opposite of "general" is not "weak," but specialized.

Today, the vast majority of productivity tools are specialized—usable in only one industry, installed on only one carrier, usable only by experts. Refer to the figure below:

From this we conclude:

General Productivity: a form of production in which "validated experience" is a public input that can be invoked across industries, across carriers, and across skill levels, so the execution floor is raised to near-expert level.

General Productivity levels:

To judge whether it is General Productivity, use three criteria:

1. Does the transfer cost approach zero?

2. Can the results be independently verified?

3. When non-experts use it, does the output floor approach expert level?

Only when all three hold simultaneously is it called General Productivity; if any one does not hold, it is still just better specialized productivity.

Historical analogy: General Power—Electricity

The goal of the Industrial Revolution was never "general mechanical intelligence," but general power.

Before the steam engine, power came from humans, animals, and waterwheels—specialized, locally sourced, and limited in capacity.

Then: steam engine → electric motor → power grid.

General power means: anywhere, at any scale, the best power can be obtained. And it was precisely this that directly brought about the Industrial Revolution and the great explosion of productivity.

The value of the electricity revolution lies not in the generator, but in the grid.

And today, every company's experience is still "self-generated electricity": accumulating experience on its own, training apprentices on its own, and when the master leaves, it is cut off.

General Productivity is to experience what general power is to energy.

The prerequisite of General Productivity: the commonization of experience

Historically, the great leaps in productivity have mainly come from the commonization of knowledge—science, education, publishing.

But experience has always been private: it belongs to one person, one workshop, one master, and cannot leave the factory gate.

Once experience can also be commonized, the returns are three multiplied terms:

Discovery cost is amortized to zero — once a piece of experience is validated, it is immediately available to the whole world;

Experience can be combined — a welder's + a materials expert's + a quality inspector's, combining into new experience;

The floor is pulled up to the teacher's ceiling.

And the productivity explosion should come from raising the floor, not pushing up the ceiling.

99% of the world's executors are ordinary people, not top experts; and top experts have a ceiling on capacity—a person only has 24 hours a day.

Pulling 99% of people to near the top yields a far greater increase in total output than making the top 10% stronger again.

The future of General Productivity: let the best productivity become everyone's confidence

We believe in people. The best machine intelligence amplifies human productivity, rather than replacing human existence. A person's value should not be determined by "how much time is sold."

We believe experience can become a public good. Printing made knowledge no longer a monopoly of the few, and the Internet made information no longer the exclusive preserve of elites. Production experience should take the same path—from individual monopoly to public infrastructure.

We believe: once experience can be commonized, productivity will usher in a great explosion.

By then, people will no longer have to make a living by "selling time"—the repetitive part is handed over, while judgment and signing remain in human hands. People can freely arrange their own time and do what they are better at and more willing to do.

But if experience always remains private, the same technology will only amplify those who already know how to do it. That is not a productivity explosion, but an explosion of productivity gaps.

Between these two things, we can only choose one. This road is long, but we are already on it.

——Intenite