Amazon has officially entered the general-purpose AI agent race with the unveiling of Nova Act, a technology that enables AI systems to take control of a web browser and perform simple actions independently. This move marks a significant step forward for the e-commerce giant, as it seeks to rival OpenAI's Operator and Anthropic's Computer Use in the rapidly evolving AI landscape.

Nova Act, developed by Amazon's San Francisco-based AGI lab, is designed to power key features of the company's upcoming Alexa+ upgrade, a generative AI-enhanced version of its popular voice assistant. Although the current version of Nova Act is labeled as a "research preview," it demonstrates Amazon's commitment to advancing AI capabilities. The Nova Act SDK, a toolkit for developers to build agent prototypes, is also now available, marking a crucial milestone in the development of AI agents that can navigate the web for users.

The Nova Act technology is Amazon's response to the growing demand for AI agents that can automate basic actions on behalf of users, such as ordering food or making reservations. According to Amazon, developers using the Nova Act SDK can create tools that enable AI agents to navigate web pages, fill out forms, or pick dates on a calendar. This level of automation has the potential to significantly enhance the capabilities of today's AI chatbots.

Amazon claims that Nova Act outperforms agents from OpenAI and Anthropic in internal tests, including the ScreenSpot Web Text evaluation, where it scored 94% compared to OpenAI's CUA (88%) and Anthropic's Claude 3.7 Sonnet (90%). However, it's worth noting that Amazon did not benchmark Nova Act using more common agent evaluations, such as WebVoyager, which may raise questions about the validity of these claims.

The Nova Act project is led by David Luan and Pieter Abbeel, former OpenAI researchers who co-founded startups Adept and Covariant, respectively, before joining Amazon last year. Luan emphasized that agents are a crucial step towards creating superintelligent AI systems, defining AGI as "an AI system that can help you do anything a human does on a computer."

Luan's team designed the Nova Act SDK to reliably automate short, simple tasks and provide developers with tools to precisely define when human intervention is necessary in an agentic workflow. This approach aims to create more reliable agentic applications, even if they are not fully autonomous. The success of Nova Act could have significant implications for Amazon's AI efforts, particularly with the upcoming release of Alexa+.

The AI agent space is becoming increasingly crowded, with major players like OpenAI, Google, and Anthropic already making significant strides. However, Amazon's entry into the market could shake things up, especially if Nova Act can overcome the reliability issues that have plagued early AI agents. It remains to be seen whether Amazon has cracked the code or if its agents will suffer from the same flaws as its competitors.

As the AI landscape continues to evolve, the success of Nova Act will be closely watched. With its wide reach and established presence in the voice assistant market, Amazon may have an advantage in popularizing AI agent technology. The coming months will provide a clearer picture of Nova Act's capabilities and its potential to revolutionize the way we interact with AI systems.