Lately, I've had a very strong feeling: it's as if I'm living in a parallel universe from many other people.
When working with the most cutting-edge models paired with Claude Code, there's a state of zero gravity. What does zero gravity mean? The speed at which an idea becomes a real product and gets validated is unimaginably fast.
Yesterday someone asked, where can I like the official website? Today I added likes. Not just likes, but comments too, and email subscriptions. All of this was completed within three hours.
The entire process was done through Claude Code plus a newly released experimental feature—it can connect directly to a Chrome browser extension, click and operate on its own, complete database integration and alignment, and obtain API keys. I only needed to handle registration and authentication at critical steps; it did everything else.
The Tsunami Has Reached Chest Level
When talking with others, I find that many people have no concept of this at all. They still understand the changes brought by AI in a very traditional way. As mentioned before, charging 500 RMB for a door-to-door installation of OpenClaw Xiaolongxia (crayfish)—this kind of absurd thing is still happening in bulk.
Anthropic founder Dario says an AI tsunami is coming right now. I completely agree. Once AI capabilities cross a certain threshold, it begins to accelerate its own evolution—the speed of connection and iteration between tools keeps increasing.
Yesterday I was still thinking, it would be great if Claude Code could integrate with Chrome. Today I discovered this feature has already been built, with an experimental version officially provided. I tried it—it's genuinely frightening.
Everything Will Be Repriced
What does zero gravity mean? The original value system will definitely be reconstructed.
I've told many people that you must use the latest models, you must spend this money—it's completely different from the free versions. Their reaction is: AI is too expensive. 2,000 RMB a month?
I say, you also spend 2,000 RMB a month on food delivery. Ridesharing costs about the same. And current prices are likely heavily subsidized by compute power—however you calculate it, those tokens you're consuming are an absolute steal. The model providers are losing money to let you use them.
But people still think it's too expensive.
When you truly feel zero gravity, all previous price anchors suddenly become illusory. Is something really worth that price? Is supply and scarcity really what we thought? All of this will change, and the speed will be explosive.
Then you leave AI and return to other fields—it feels like crawling at a turtle's pace.
Speed Is Saving Value
I've developed new thoughts about competition.
Before, in a very small market, the first instinct was to think about competition. Now I don't think that's the issue. When you're facing trillions of dollars in value creation, time and speed are saving value itself.
How to understand this? If we replace inefficient production methods with more efficient ones and popularize them faster, we are saving societal value.
Do the math. Suppose a service originally costs 2,000 RMB per year, and the new method drops it to 100 RMB—a 95% savings. Covering 100 million users, the total savings would be roughly 190 billion.
If adoption were completed instantly, all 190 billion would be saved. But in reality, it rolls out linearly—taking a full year slowly turns the rectangle into a triangle, leaving only half, less than 100 billion.
The difference in speed directly determines how much value is saved.
No Competitors
So in this environment, we aren't competing against anyone. Everyone is a participant.
This market is too large. Too large to naively believe that one person or a small team can capture 200 billion in total value within a single year. Who will do it? Everyone, together, building out this market.
Wang Xing (founder of Meituan) once talked about Infinite Games and Finite Games. Under the wave of new productive forces emerging, it's like the Western Expansion—from existing stock to new increments, full of possibility. The first priority isn't thinking about competition, but who can make the pie grow faster. Those who move faster naturally create more value and rightfully deserve greater returns.
Stop and Look
Traditional value systems will soon be overturned. The tsunami has already risen to chest level, yet many people haven't even felt it.
I've been urging people: at least download it and experience what the most cutting-edge models can do right now.
I've demonstrated this to others. One sentence—"Help me install OpenClaw locally." The AI executes on its own, overcoming all obstacles along the way: compiling when compilation is needed, resolving dependencies when dependencies are needed. It ran continuously for about half an hour, then delivered it, saying it's ready to use.
Everyone goes silent after watching.
This official website itself is the most representative example. I'm not a technical person. Someone gave feedback wanting likes, and they appeared the next day. Then not just likes, but comments too. Not just comments, but subscriptions too. Two or three hours—from thought to completion. The entire process from idea to being seen took only two days.
What's the cost behind it? A few thousand RMB per month in model subscription fees.
The tsunami is already here. Those still working in the old ways really need to stop and look at what's happening. The software bidding system may be the first industry to collapse.