CEO of a top Silicon Valley AI company: The next real money-making opportunity is here.

A few days ago, Anthropic held a developer conference in Bangalore.

There were two people on the stage: CEO Dario Amodei and CTO Rahul Patil.

Dario is the father of Claude and the leader of the "AI safety faction". Rahul is a master of infrastructure construction. He personally participated in building three of the world's four major cloud platforms and also worked at Stripe on the underlying system that supports global financial flows.

One of them is in charge of "thinking", and the other is in charge of "executing".

You may not like what they said on the stage. But I think the less you like it, the more worth reading it carefully.

01 Pour a bucket of cold water first

The first thing Dario did when he took the stage was to tell the truth to the entrepreneurs doing AI applications in the audience:

"If you only create a pure software UI interface, your moat is almost non - existent."

This statement hit a lot of people.

In the past two years, the story of making money by "wrapping large models" has spread widely. Find an API, wrap an interface, and sell subscriptions. It seems that this is how AI entrepreneurship should be done.

But Dario said that this path is getting narrower and narrower. The reason is simple: the capabilities of large models are constantly "trickling down". The functions that require your careful tuning today may be directly achieved by the new models in a few months. The wall you painstakingly built is made of paper.

So what is the real moat?

The answer he gave is just four words: The atomic world.

That is, the things that truly exist in the physical world - materials, molecules, the human body, and drugs.

Software can be infinitely replicated, but no one can replicate the data of a drug that has gone through three phases of clinical trials.

02 The biggest opportunity lies in the "difficult" areas

Dario said that the direction that excites him the most personally is biology and medicine.

It may seem that this has nothing to do with most AI entrepreneurs.

But he made a very memorable prediction:

"Imagine that the medical innovations of the past 100 years will explode in the next 5 to 10 years."

Why medicine? Precisely because it is difficult.

You need to have professional knowledge, pass FDA approval, conduct clinical trials, and connect with real hospitals and patients.

These "difficulties" are exactly where the opportunities lie - "It is precisely because of these difficulties that you can build a long - lasting business."

Moreover, the reasoning ability of AI makes this a real possibility.

Claude is not only good at writing code but also at reasoning. An AI that can derive physical formulas from first principles can also derive molecular structures and understand protein folding.

This is not just memorizing literature but truly "thinking like a scientist".

03 An eye - opening metaphor

Dario quoted a saying from hockey legend Wayne Gretzky:

"Skate to where the puck is going, not where it is."

He said that this is the most important advice he gives to all entrepreneurs.

Many people complain about the current flaws when using AI: this model is not smart enough, and that function is not good enough. Then they think about problems and design products based on the existing limitations of capabilities.

Dario said that this is the biggest mistake.

The capabilities of AI models will have a step - by - step leap every few months. If you design based on the existing capabilities today, by the time your product is made, the model will have already been upgraded, and what you have made will seem conservative and outdated.

Rahul put it more directly: "If an application only seems 'barely feasible' today, the new models in a few months will make it 'absolutely feasible'."

Therefore, those who dare to bet on the future direction when the technology is not yet perfect are the ones who can finally reap the benefits.

04 Will you be a spectator or a builder?

What has been said above may seem like something for entrepreneurs, but there is the same logic for every ordinary person:

Understand the direction and then use AI to amplify yourself. How can you seize the opportunities of your era?

First, find your irreplaceability.

What AI has the most difficulty replicating is not skills but the real experience and judgment you have accumulated in a certain field. Those who have worked in sales, medicine, or teaching - your industry understanding combined with AI's execution ability is the real one - two punch.

Then, use AI proficiently.

Make it a part of your daily work. Those who have used it for 100 days and those who have used it for 1 day are definitely not in the same position.

Finally, dare to have great ambitions. "Dream the biggest dreams and build the seemingly impossible."

With the empowerment of AI, you today are different from you yesterday. However, the lack of imagination limits the boundaries of our lives.

Rahul told a story from more than 100 years ago at the end of his speech.

In 1899, someone said: "Everything that can be invented has been invented."

As a result, a few years later, the director of the US Patent Office left another sentence: "All past inventions will seem insignificant compared to what is to come."

Today, we are standing at the position of 1899 again.

The only difference is whether you are the one who ignores the changes or the one who has already started to bet on the future.

This article is from the WeChat official account "Kelly Peng", author: Kelly Peng. Republished by 36Kr with authorization.