NeuralAgent 2.5 Launches with Voice Mode and Watch & Learn Features

Khaled and the Neural team have released NeuralAgent 2.5, described as their most significant update to date. NeuralAgent is an AI agent that displays a user's screen and autonomously controls their computer by opening applications, clicking buttons, managing files, and executing tasks across the entire system. Users can describe tasks in plain English and observe the agent complete them automatically.

The new version introduces Voice Mode, which allows users to communicate with NeuralAgent verbally while the agent responds and controls the computer simultaneously. Khaled highlighted this feature as the closest real-world implementation to the fictional AI assistant Jarvis. The second major feature, Watch & Learn, enables users to demonstrate a task to the agent once, after which the agent saves the procedure and can repeat it on demand. This addresses the practical limitation of repeatedly explaining identical tasks to an AI system.

According to the announcement, NeuralAgent is used by tens of thousands of people globally, including Fortune 500 enterprises. The developers emphasize that the release bridges a significant gap in AI assistant capabilities: the ability to reliably complete complex, real-world tasks across multiple applications without constant supervision. The team remains focused on handling multi-step workflows and ensuring consistent task completion.