AI applications are moving beyond the chat box, and two recent updates show where that shift is heading. Wabi is pivoting toward a messaging experience that can make apps and help users complete tasks, while ChatGPT is bringing voice-based agentic features to its mobile app.
The direction is clear: users want AI that can talk, work, and present the right interface at the right moment. Instead of forcing every task through a single stream of messages, these products are building toward assistants that combine conversation with action.
Wabi Wants Messaging to Become a Working Interface
Wabi is undergoing a slight pivot to become an AI messenger capable of making apps and helping users get things done. Wabi announced its pivot to a messaging experience on September 28, 2026, and Sarah Perez published an article about Wabi on September 29, 2026.
Wabi 2.0 is described as a “personal agent that does stuff for you and builds the interface you need in the moment.” That idea places the app itself inside the conversation, allowing the agent to create a useful interface when a request needs more than a written answer.
The goal is to compete more directly with AI agents that combine conversations and productivity tasks in a single interface. Meta has AI agents like Muse and Instinct, giving Wabi a clear group of products to challenge as it reshapes its experience.
Eugenia Kuyda, the founder of Wabi who previously founded Replika, has argued that chat-only tools reach a limit when users need to manage ongoing work. As Kudya put it:
“The more you use a chat-only agent, the worse the experience gets. Your history and ongoing tasks end up buried in one endless scroll. People don’t just want to type-type-type; They want to tap-tap-tap, scroll, and look. They love what agents can do, but they also like software.”
That criticism points to the central promise of Wabi 2.0: a conversation can remain the starting point, but the result does not have to stay trapped in text. Kudya described the product direction this way:
“The idea is that as a user asks the agent to perform a task, it could create a user interface on demand to help complete whatever the request may be.”
ChatGPT Brings Agentic Voice Features to Phones
OpenAI announced voice-based features for the ChatGPT mobile app on September 23, 2026. The update gives Plus and Pro users access to the Work tab on their phone, where they can create a document, draft an email, or summarize Slack messages.
These tools connect voice conversations with productivity tasks, letting users start with spoken instructions instead of typing every request. ChatGPT’s voice conversations will also have richer text output, and users will be able to switch between text and voice easily.
The mobile experience is built around movement between devices and settings. Users can start a conversation on the go and resume it on desktop, creating a path from a quick phone interaction to continued work on a larger screen.
OpenAI launched GPT-Live in July 2026, and the new mobile voice features extend that direction into agentic work. The assistant is not limited to answering a question aloud; Plus and Pro users can use the Work tab to turn requests into documents, emails, and Slack summaries.
- Create a document from the Work tab on a phone.
- Draft an email through the mobile experience.
- Summarize Slack messages.
- Start a voice conversation on the go.
- Resume the conversation on desktop.
- Switch between text and voice with richer text output.
The Next AI Interface May Change With the Task
Wabi and ChatGPT are taking different routes toward a shared idea: AI agents need more ways to help users than an endless scroll of replies. Wabi is focused on creating an interface on demand, while OpenAI is combining voice, text, mobile access, and productivity actions inside ChatGPT.
That shift could make the interface itself part of the agent’s response. A user might ask for help, speak through the first step, review richer text output, and then use a generated workspace or productivity tool to finish the task. The exact experience differs between Wabi and ChatGPT, but both updates place action alongside conversation.
The timing also creates a busy moment for AI applications. Wabi’s announcement came on September 28, 2026, followed by the September 29 article about its pivot, while OpenAI’s announcement arrived on September 23, 2026. Each update pushes the same question into view: should an AI assistant only answer, or should it build the tools needed to complete the request?
There are also two upcoming dates tied to the surrounding technology activity. October 2, 2026, is the last day to demo a breakthrough to 10,000+ tech leaders. Disrupt ticket savings of up to $200 plus 50% off a second ticket end September 25, 2026, at 11:59 p.m. PT.
The next phase of AI assistants will be measured by what happens after the conversation begins. Wabi wants to create the interface users need in the moment, while ChatGPT is giving mobile users a voice-driven route into documents, emails, and Slack summaries. The winning products may be the ones that know when to talk, when to listen, and when to turn a request into software.
Based on
