A new study published this week reveals that OpenAI is developing a ChatGPT agent capable of performing various tasks on smartphones. The agent will be able to receive commands and perform actions like clicks, cursor movements, and text input, similar to how humans interact with apps. This development aligns with Google’s previous efforts to create a more advanced assistant, as demonstrated with the release of the new Google Assistant on the Pixel 4 in 2019.
During the I/O 2019 event, Google showcased its next-generation assistant, which aimed to revolutionize the way users interact with their smartphones. The assistant boasted near-zero latency, allowing users to control their phones with their voice and seamlessly multitask between apps. It even had the ability to coordinate tasks across different applications, demonstrating a high level of multitasking and complex action completion.
However, the new Assistant introduced in 2019 had limitations. Users had to stick to specific phrases and commands, and the assistant did not have the capability to adapt to different conditions. While Google promised improved app integration, some of the showcased features no longer work due to changes in the app landscape.
Therefore, OpenAI’s ChatGPT agent and Google’s previous research on enabling conversational interactions with mobile UIs using large-scale language models present a potential solution. OpenAI’s agent, along with Google’s research, could enhance the capabilities of the new Google Assistant, allowing it to understand commands more naturally and adapt to set conditions.
The Gemini AI agent for Android devices is seen as a natural evolution of Google’s initial efforts to create a comprehensive assistant. It offers new ways to utilize smartphones, such as transcribing and sending messages through voice input in the Gboard Assistant.
Overall, these developments highlight the ongoing efforts in the industry to create AI agents that can perform tasks on smartphones, bridging the gap between user commands and physical actions. Google has the opportunity to prioritize and enhance its Assistant capabilities, empowering it to lead in this area rather than playing catch-up. With the potential for improved app integration and more natural language understanding, the future of AI assistants on smartphones looks promising.
