OpenAI has announced its AI agent tool, called Operator, as a research preview as of Thursday, but the launch isn’t without its minor hiccups.
The artificial intelligence brand showcased features of the new tool in an online demo, explaining that Operator is a Computer Using Agent (CUA) based on the GPT-4o model, which enables multi-modal functions, such as the ability to search the web and being able to understand the reasoning of the search results.
However, those who have gotten a chance to test the tool have noted critiques, such slow responsiveness in comparison to the demos, and hallucinations akin to the standard ChatGPT chatbot, according to Quartz.
Similarly, Bezingza reported that some X users’ complaints got the attention of OpenAI CEO, Sam Altman. According to the publication, an X user shared on the platform issues with Operation’s interactions with a news website. The CEO reportedly responded, promising a prompt fix of the issue. However, what the user experienced could potentially be a hallucination.
While many are fascinated by the features of Operator showcased in OpenAI’s launch demo, price remains a big deterrent from experimenting with the AI agent. The tool is available under OpenAI’s $200 per month ChatGPT Pro tier, making it a rather exclusive experience.
BGR writer Chris Smith noted being a ChatGPT Plus subscriber, but said he could not justify paying $200 per month to access the Operator tool. However, OpenAI is expected to bring Operator to its ChatGPT Plus, Team, and Enterprise tiers at some point.
One of users’ biggest complaints is that the Operator is currently available only in the U.S. European users have expressed disdain that they cannot access the tool.
ComputerWorld also noted that AI agents as a whole can potentially introduce unique safety risks, such as using automated ecosystems to launch traffic attacks or bypass CAPTCHA codes. While OpenAI insists it has a security on lock on Operator, the publication spoke with researchers that noted the technology could clash with search engines, such as Google, that have their own data processing strategies in place for its own purposes.
