OpenAI Engineer Guinness Chen Announces Full Launch of ChatGPT Voice and Apologizes for Delay
OpenAI Confirms Full Launch of ChatGPT Voice
OpenAI engineer Guinness Chen announced that ChatGPT Voice has been fully launched and apologized for the delay.
In the launch video, a joke was made about the "function switch being completely misconfigured," which turned out to be true as the actual misconfiguration caused the delay, inadvertently resulting in a documentary.
The feature launch is event-driven, allowing users and developers to switch to voice interaction mode in real-time. Funding and attention are flowing towards AI applications that support full-duplex voice, benefiting developers in voice-first scenarios while putting pressure on pure text interaction products.
Source: Public Information
ABAB AI Insight
Guinness Chen is responsible for OpenAI's Codex and ChatGPT Voice, having previously demonstrated the integration of voice with screen sharing and task execution on desktop, indicating that the company is upgrading voice from an auxiliary feature to a core interaction layer.
On the capital front, OpenAI continues to invest R&D resources into real-time voice models and full-duplex architecture, motivated by the desire to reduce user input friction and extend daily usage time, while locking in more paid subscriptions and API calls through the voice interface.
Similar cases can be seen with the early full rollout of Google Assistant and Amazon Alexa, which are currently in an expansion phase transitioning from experimental voice modes to default interaction methods.
This essentially represents a technological substitution, where the mechanism is that real-time voice reduces cognitive load, transforming AI from a tool into a continuous conversational partner, accelerating the marginalization of text interfaces.
ABAB News · Cognitive Laws
Jokes often come true
Function switches determine product fate
The more natural the voice, the more marginal the text