A Hugging Face team has launched an “agent” that uses the computer that uses the free use computer. But be careful: it is quite slow and occasionally made mistakes.
Hugging the face of the face, called Open computer agentIt can be accessed through the web and can use a virtual Linux machine with several applications, including Firefox. Similar to OpenAI operatorYou can ask Open Computer Agent to complete a task, say, “Use Google Maps to find the HQ of the face hugged in Paris”, and feels as the agent opens the necessary programs and discovers the required steps.
The open computer agent can handle simple requests well enough. But the most complicated, such as looking for flights, they took it out in the TechCrunch tests. Open Computer Agent is also often executed in captcha tests that cannot be resolved.
You will also have to wait in a virtual tail to use an open computer agent, a second tail up to minutes, depending on the demand.
Of course, the objective of the clamp team was not to build a state -of -the -art computer use agent. Rather, they wanted to show that open AI models are becoming more capable, and cheaper to execute in the cloud infrastructure.
“As the vision models become more capable, they become able to feed complex agent workflows,” Aymeric Roucher, member of the Hugging Face agents team, wrote in a publication about X. “[Some of these models] Incorporated grounding support, that is [the] Ability to locate any element in an image for its coordinates, [and] thus [can] Click any item [in a virtual machine]. “
While it is far from perfect, agent technology is attracting a growing investment as companies seek to adopt it to boost productivity. According to a recent KPMG survey65% of companies are experiencing with AI agents. Market and market projects that the AI agent segment will grow from $ 7.84 billion in 2025 to $ 52.62 billion by 2030.
Techcrunch event
Berkeley, ca.
|
June 5