OpenAI is building AI agents for everything. Will everyone use them?

OpenAI is developing artificial intelligence agents capable of handling complex tasks across various digital workflows, ranging from accounting to medicine. To achieve this vision, OpenAI is building tools that allow non-technical users to interact with large language models (LLMs) in a more intuitive and seamless manner.

Andrew Ambrosino, the lead engineer for OpenAI's desktop app, explained that giving LLMs access to sensitive data like emails and chat logs is a risk. However, he is willing to take the personal hit if it helps advance the technology. The company's ChatGPT Work, released last month, is accessible on the company's lowest subscription tier and aims to bring AI agents to a broader audience, not just software engineers.

OpenAI's marketing suggests that its goal is to make AI a tool for turning ideas into reality, not just answering questions. By creating agents that can handle multi-step projects autonomously, OpenAI aims to make AI accessible to a wider range of users. However, reaching new professional markets like law and sales is crucial for the company's commercial success.

The challenge lies in making AI agents work for users who aren't software engineers, as traditional interfaces like command-line interfaces (CLIs) are no longer user-friendly. OpenAI is addressing this by developing agentic products that can interface with existing tools and websites, abstracting away the technical complexities.

Industry analysts see this as one of the major challenges facing OpenAI and its competitors. To achieve widespread adoption, OpenAI needs to find a balance between providing power users with the functionality they need and making the technology accessible to mainstream users.