JAKARTA - The world of technology is being enlivened by the emergence of a new artificial intelligence framework called OpenClaw which is said to be developing at an incredible speed. Nvidia's chief executive, Jensen Huang, even called it one of the most important software releases in the history of the technology industry.
Speaking at a conference hosted by Morgan Stanley, Huang said OpenClaw's adoption has skyrocketed far beyond the growth curve of classic open source projects like Linux.
"OpenClaw achieved an adoption rate that Linux took 30 years to achieve in just three weeks," Huang said.
OpenClaw, previously known as MoltBot and ClawdBot, is a framework for building and coordinating "AI agents". This concept marks a new evolution in the use of artificial intelligence.
The first wave of AI focused on conversation and answering questions. But a new phase called "agentic AI" places AI as an autonomous task executor. Instead of just answering questions, the system can now be instructed to create, build, or run complex work.
In practice, these AI agents can independently browse the internet, fix software bugs, and handle technical workflows that previously required human intervention.
According to Huang, the shift from a simple question-answering model to a system that performs active tasks creates a new phenomenon he calls the "compute vacuum".
Computing demand increases drastically as AI agents work continuously in the background. A typical AI prompt uses relatively little computing resources. However, AI agents working on solving problems can consume up to 1,000 times more tokens.
In some cases, agents that run continuously for 24 hours to monitor and optimize a company's systems can even consume tokens up to one million times larger than a typical AI conversation.
This surge in computing needs is a big opportunity for hardware providers. But at the same time, it also challenges the current technology infrastructure.
NVIDIA's AI hardware such as the Hopper and Blackwell chips were initially designed primarily to train large AI models. Now the industry is beginning to shift towards systems capable of handling longer and more complex AI agent workloads.
Huang also said that OpenClaw's growth is almost vertical when viewed in a technology adoption graph. According to him, the main cause of the spike is because this technology makes AI much more practical for both daily work and professional software engineering.
Within NVIDIA itself, the OpenClaw agent has already been used to develop software as well as build various internal tools.
Huang believes the use of AI agents will become the new standard for modern companies. Companies that are faster to adopt this technology are expected to experience a significant productivity boost, while organizations that are late in integrating it risk falling far behind.
To keep up with this surge in computing needs, NVIDIA is also preparing the next generation of hardware with the code name Vera Rubin. This system is designed to increase the memory capacity and long-context processing capabilities needed by AI agents.
Behind all the tech jargon, this change is actually simple but radical. AI is evolving from being merely a "question answering machine" to a "worker machine". As software begins to act like digital employees who never sleep, the computing world will require energy, chips, and infrastructure on a much larger scale than before.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
