The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.

In 1987, Apple’s vision of computing foresaw work happening within tools rather than between them, a concept now materializing through agentic AI. Instead of opening separate programs to analyze data or create charts, users can now ask an AI agent to perform these tasks directly within platforms like Slack or WhatsApp. Open-source agents such as OpenClaw and Hermes execute instructions in these environments, while tools like Claude Code can generate custom interfaces tailored to individual needs, such as adjusting text size for accessibility or providing voice-based interactions. This evolution transforms software from a destination into a capability that adapts to the user’s context and workflow.

The shift reduces reliance on traditional menus and dashboards, as AI agents interact directly with underlying data and software logic. For example, a manager might approve an expense report from within a chat window without opening the dedicated expense app, or an analyst could pull sales data into a presentation without logging into a separate dashboard. The user’s primary interface becomes the agent itself, which handles execution while the user focuses on oversight, review, and governance. This mirrors long-standing visions of ubiquitous computing, where technology fades into the background, and the interface melts away, leaving only the completed task or decision.

Voice interaction is emerging as a key component of this new interface paradigm. Tools like Salesforce’s Voice to Form allow users, such as water treatment inspectors at Workdry Group, to dictate compliance surveys onsite in real time, reducing post-visit documentation from hours to minutes. The next phase involves continuous voice processing, enabling more natural, conversational interactions where the AI listens and responds in real time, much like a human colleague. Ambient intelligence further enhances this by quietly analyzing calendar entries, messages, and other digital traces to surface relevant information across devices like phones, smartwatches, or future wearables.

The future of AI interfaces may involve a mix of new devices—such as smart glasses or earbuds—and existing platforms like smartphones, which remain central due to their ubiquity. Operating systems and hardware are also evolving to support this shift, with companies like Nvidia introducing chips designed to run AI agents locally. Salesforce’s chief scientist, Silvio Savarese, describes this as 'ambient intelligence,' where software proactively understands context and 'reaches you before you reach for it.' This could mean a sales rep receiving critical customer details mid-call or an operations manager being alerted to the one important signal amid a flood of noise.