Anthropic has launched a new “computer-use” AI tool which is capable of automating your computer by handling the mouse cursor.
In a race to build competitive AI tools and software, Amazon-backed Anthropic takes a step forward by developing an AI tool that can automate mouse clicks in computers. Likely as ‘on-screen awareness’ by Apple Intelligence, this AI tool by Anthropic understands the screen and indeed controls the mouse too.
Anthropic released a pair of updated artificial intelligence models recently, along with this new capability to autonomously perform computer tasks and save users keystrokes.
Computer use AI feature by Anthropic
Dubbed as “computer use”, the AI feature can tell “where to move the mouse, where to click, what to type, in order to do quite complicated tasks,” Anthropic’s Chief Science Officer Jared Kaplan said in an interview.
The computer-use AI feature is tailored to software developers and represents a move toward AI agents, programs that require little human intervention to carry out multi-step actions.
In an example, Anthropic demonstrated the use of feature that entailed coding a basic website and in another, the feature was seen using various apps including Google Search and Apple Maps to plan a sunrise outing automatically.
The feature, if combined with Anthropic’s generative-AI model Claude can automate the entire process for which an entry-level employee is hired to do.
Anthropic’s Claude
Anthropic’s ChatGPT-rival “Claude” excels in handling vast volumes of text, data-heavy files and in analysis of tasks, when compared to the fore-runner OpenAI’s ChatGPT.
Further, Anthropic has launched two new versions of Claude – Sonnet, and Haiku, updating its family of AI models.
Clade 3.5 Sonnet, the mid-tier model, comes with this computer-use feature and safeguards to prevent is application toward spam, fraud and election-related misuse.
Whereas the new cheapest AI – Claude 3.5 Haiku model can generate code in an ‘almost comparable’ manner to the version of Sonnet.
The AI models when put to use with computer-use feature can self-correct and retry tasks when it encountered obstacles, said Anthropic.
The approach that Claude has taken is completely different from that of currently existing AI models. Up until now, LLM developers have made tools-fit-the-model approach, producing custom environments where AIs use specially-designed tools to complete various tasks. Now, Claude have followed the model-fit-the-tools approach where AI can fit into computer environment we all use every day.
There’s still a lot to do. Even though Claude is the current state of the art, Claude’s computer use remains slow and often error-prone. There are many actions that people routinely do with computers (dragging, zooming, and so on) that Claude can’t yet attempt. There are also several mistakes Claude has done, Anthropic reported. However, the firm has clearly defined the track for its destination and is optimistic on achieving it.
(For more such interesting informational, technology and innovation stuffs, keep reading The Inner Detail).
Kindly add ‘The Inner Detail’ to your Google News Feed by following us!
