The race to develop advanced artificial intelligence (AI) agents — autonomous AI systems designed to perform complex tasks and interact with the digital world on behalf of users — is heating up. For example, San Francisco-based Zeta Labs raised $2.9 million in a pre-seed funding round.
“At Zeta Labs, we aim to enable AI models to interact with the digital world around them. The development of our AWA-1 model allows JACE the ability to control a browser, essentially giving the AI assistant the digital equivalent of arms and legs,” Peter Albert, co-founder of Zeta Labs, said in a news release. “You can’t ask any current chatbot on the market to book a trip, pay an invoice or set up a job post. With JACE, you can.”
Zeta Labs’ AI agent, JACE, aims to bridge the gap between AI and the web by going beyond the capabilities of traditional AI chatbots. While most chatbots focus on text generation, JACE is designed to navigate and use digital tools to assist users in various tasks, from making reservations to setting up business processes.
According to Zeta Labs, by teaching JACE to control a browser, they have created a system capable of emulating a substantial portion of day-to-day office work. To showcase JACE’s capabilities beyond the traditional AI-assistant role, the company reportedly tested the agent by asking it to create a company independently. Starting with a simple prompt, JACE guided the user through creating a business plan, registering the business as a math tutoring company, finding its first client, and generating its first revenue within two weeks.
JACE demonstrated impressive performance in an internal benchmark test conducted by Zeta Labs, the company reported. The test focused on common web-based tasks such as navigating the ordering process for pizza delivery or finding two flats in London that meet specific criteria. The company claims that JACE performed best when using Zeta Labs’ proprietary AWA-1 model, achieving an 89% task completion success rate, compared to a 68% success rate when using GPT-4o and a 25% success rate when using an open-source web agent with GPT-4o.
“In its current form, I see JACE as a meta-aggregator for web interfaces. Why learn to navigate all the different user interfaces when you can have a universal UI for everything?” Fryderyk Wiatrowski, co-founder of Zeta Labs, said in the news release.
“With JACE, you simply say what you need, and if you forget, it will ask you all the necessary questions up front. Imagine a universal service desk with a consultant who knows you personally and never needs to sleep. JACE will be that for you, just sitting there, ready to serve you at all times.”
However, the company must still prove its AI agent’s real-world utility and reliability. “It’s not clear cut, though it does look similar to agents like Perplexity,” Muddu Sudhakar, the co-founder and CEO of Aisera, a provider of generative AI for enterprise, told PYMNTS. “Zeta Labs says it performs better than GPT-4o, but this is not based on third-party evaluations.”
As AI technology advances, new AI agents emerge, potentially transforming how businesses interact with customers. PYMNTS previously reported that Altera is developing these AI agents by drawing inspiration from neuroscience. “The brain is the only example of a high-functioning agent we have,” said Altera Co-Founder Robert Yang.
Despite the early stage of the technology, experts believe that autonomous AI agents could significantly improve productivity across various industries. “Agents are a red-hot category. It’s the next phase in the progression of generative AI — from traditional chatbot communications to systems that can carry out actions, reason and plan,” Sudhakar explained.
Aisera, for example, has been developing AI agents focused on enterprises, integrating them with common ERP and CRM systems, along with other corporate platforms. “If an employee has a question about benefits, our agent can both answer it and also take actions. This could be to change a healthcare policy, add a beneficiary, and so on,” Sudhakar added.
The funding round for Zeta Labs, led by investors Daniel Gross and Nat Friedman, along with participation from Earlybird VC, Kaya VC, AI Grant, and others, demonstrates the growing interest in the potential of AI agents. “It’s another data point of interest for savvy investors in the category. Expect many other deals. There are many startups popping up that are developing agents,” Sudhakar said.
As the AI industry continues to evolve, companies like Zeta Labs and Aisera are pushing the boundaries of what AI can do. While the development of autonomous AI agents holds promise, it remains to be seen how these systems will integrate into our daily lives and whether they can live up to the hype surrounding them.
We’re always on the lookout for opportunities to partner with innovators and disruptors.
Learn More