Lobster Skill is just the appetizer; OpenClaw is recreating the eve of the iPhone's breakout.

“Lobster” OpenClaw has become completely popular in China, so much so that even Pony Ma shared related news about Tencent’s free open-source AI assistant OpenClaw on his Moments, commenting, “Didn’t expect it to be this popular.”

During the National People’s Congress, Chinese Academy of Engineering academician Gao Wen also mentioned this phenomenon, saying, “Everyone is so anxious now, afraid they won’t be able to raise lobsters.”

But now that “lobsters” are here, what exactly are people doing with them? A fairly normal and ideal example might look like this:

“About a week ago, I’ve been using a digital assistant that knows my name, understands my morning routines, knows how I like to use Notion and Todoist, and can control Spotify, my Sonos speakers, Philips Hue lights, as well as my Gmail. It runs on Anthropic’s Claude Opus 4.5 model, but I communicate with it via Telegram. I named this assistant Navi, and Navi can even receive my voice messages and respond using the latest ElevenLabs text-to-speech model. Oh, and I haven’t mentioned that Navi can improve itself through new features, and it’s running on my own M4 Mac Mini server.”

The author of this case study also mentioned that he has burned through 180 million tokens on the Anthropic API, possibly spending around $2,000 to “raise a lobster.”

It might seem that the cost of “raising lobsters” isn’t low, and what they can do doesn’t seem particularly grand—more like a “XX spirit” that can communicate with humans in a human-like way and help automate more tasks. In reality, this is the current role of “lobsters”—an “AI assistant.”

If we look at the top 100 installations on ClawHub and categorize them roughly, we’ll see that using large models for these tasks is often an overkill:

• Information retrieval: Search, extract, integrate, and summarize information from various sources (external links, local files, APIs). Use cases include AI-optimized Google or Baidu searches, daily weather reports from “lobsters,” real-time Bitcoin prices, etc.

• Productivity (workflow automation): Handle emails, Notion, Github, Obsidian, Slack, and enable cross-platform task automation, simplifying workflows—one entry point for multiple platforms.

• Developer tools: Professional tools for developers and technical users, offering code management, API interaction, server management, etc. Improving development efficiency through automation of code, testing, deployment—like command-line interactions with GitHub, issue and PR management, CI runs, and advanced queries.

• Content creation: Use AI’s generative capabilities to create or edit text, images, audio, and other multimedia content.

• IoT control: Connect and control smart home devices, audio systems, and other smart hardware—like scheduling curtains and lights to turn on/off at specific times.

Overall, the explosive popularity of “lobsters” isn’t because they excel at these functions, but because they act like a “secretary”—comprehensive enough. Compared to most users who might simply treat an AI tool as a search engine or automatic image editing software, “lobsters” allow people to give multiple tasks via chat apps like Telegram, just like talking to a boss. This novelty amplifies through word of mouth, creating an unprecedented experience of AI entering daily life.

We can even view the current seemingly idle stage of “lobsters” more optimistically. In the early days of the iPhone, we only used it to play games like Balance Ball, Angry Birds, Fruit Ninja—“demonstrating what touchscreens can do.” In terms of content and fun, these games weren’t even as engaging as Nokia’s Java games. But now, young people play Honor of Kings, Delta Force, and many only play mobile games, not PC games.

If we turn our attention to today’s cryptocurrency market, “lobsters” could once again significantly lower the barrier for the public to learn about and participate in crypto, effectively addressing widespread investment needs.

Of course, this doesn’t mean trading memes or issuing tokens with “lobsters.” Today, the assets tradable on-chain are becoming increasingly diverse—US stocks, crude oil, gold, Pokémon cards… We can trade all of these in a decentralized, 24/7, no-barrier manner. The trading volume is substantial: on February 6, Hyperliquid’s on-chain Perp DEX and Trade.xyz, mainly dealing with US stocks, reached a 24-hour trading volume of $5.45 billion, setting a new record.

In an era of abundant information, what often blocks us from catching new investment opportunities is “not having the right access.” For example, recently, memory prices surged—everyone knew this, but for non-Koreans, directly buying SK Hynix stock was complicated. Opening accounts, settling funds—these obstacles delayed immediate investment actions.

But if the process becomes:

• Giving “lobster” a wallet

• Using a credit card to buy stablecoins and fund the “lobster” wallet

• Telling “lobster” the specific assets you want to invest in

• “Lobster” executing buy/sell orders on-chain

And all this is done as simply as chatting with friends, it would be an explosive growth opportunity for “lobsters” and crypto.

We also have prediction markets, so we can imagine more scenarios. For example, chatting with a driver during a taxi ride. The driver says he thinks candidate A will win the next US presidential election, but you believe it’s B. When in disagreement, you give a voice-to-text command to your “lobster”—“Help me bet $100 that B will win.”

Your “lobster” understands your intent, automatically finds the most liquid prediction market, and places the order. The driver quickly follows, using voice commands through the car’s system, also betting $100 on A winning, via “lobster.”

Even “lobsters” might need to implement restrictions on minors’ consumption—to prevent kids from impulsively spending on Pokémon cards in the chain’s card market while showing off their collections.

If the pump.fun meme coin craze sparked the “everything tokenized” attention economy 1.0, then “lobsters”—a new paradigm where ordinary people can easily use AI—could become version 2.0. It can find all the assets and channels we want to invest in instantly on-chain and execute according to our intentions. Moreover, it will expand the on-chain ecosystem from investment to consumption, truly realizing the long-sought Mass Adoption of blockchain.

The future is happening.