My thoughts about ClawdBot
The hottest topic in the tech world lately is undoubtedly ClawdBot. While many people have been using this little tool, I eventually decided to try it out myself.
To be honest, I initially thought it was just a toy for developers that couldn’t generate any real productive value. However, after testing it for a while, I have completely changed my mind.
What is ClawdBot?
ClawdBot is actually very similar to Manus, which went viral earlier. Both are AI assistants with autonomous capabilities.
Compared to conventional large language models—such as Google’s Gemini or OpenAI’s ChatGPT—these assistants are not satisfied with being just a wise mind. Instead, they have equipped themselves with robust hands and feet, giving them the ability to take independent action.
To put it more simply, ClawdBot is like a llm that can execute tasks automatically. You don’t need to manually provide it with context; it can control your device on its own, view necessary content, and modify the files you want to change.
A major difference is that ClawdBot can also arrange proactive plans for the future, changing the previous situation where large models could only wait passively for instructions.
Additionally, ClawdBot can access a wide variety of skills, allowing it to visit and operate third-party tools such as:
- Apple Reminders
- TickTick
- Things
As more surrounding software begins to provide support, ClawdBot still has a lot of potential to be uncovered.
My stack to run ClawdBot
I have a spare M1 Mac mini with only 8GB of RAM, but my testing shows that it is more than enough. I deployed ClawdBot directly onto this Mac mini, and the entire system is running perfectly.
After several days of testing, I’ve found that ClawdBot’s resource consumption is quite low; standard hardware can easily support its operation. Interestingly, the popularity of ClawdBot has reportedly led to a surge in Mac mini sales across various e-commerce platforms.
This brings me to a personal theory: the future of AI will be dominated by companies that, like Apple, excel at integrating hardware and software. A standalone LLM does not fully unleash the power of artificial intelligence.
If a manufacturer can produce a beautifully designed device with powerful performance—paired with the right software—that kind of product will undoubtedly lead the future consumer market.
Final thoughts
Many people in our generation are fans of Marvel’s Iron Man and his partner, Jarvis. As artificial intelligence continues to improve and grow more powerful, owning a personal Jarvis no longer seems like an unattainable dream.
We are currently at a critical turning point where social and productive relations are undergoing rapid development and massive transformation. I believe that in the future, there will be even more and better AI projects capable of integrating AI capabilities into our daily lives.
Ultimately, this wave of AI will sweep from the software world into our reality. As a result, humanity’s vision for the future may need to undergo a significant shift.
