OpenAI Temporarily Discontinues Sky Voice Option
In a statement provided to NPR by the actor’s publicist, Scarlett Johansson revealed that OpenAI had contacted and asked her to provide a voice for ChatGPT 4.0 in September 2023, but after careful consideration she declined for personal reasons. In a statement provided to NPR by the actor’s publicist, Scarlett Johansson revealed that OpenAI had contacted her in September 2023 to offer her voice for ChatGPT 4.0, but that she had thoughtfully declined for personal reasons, and that Scarlett Johansson had asked OpenAI to disclose the process of creating the voice option in light of the overly similar tone in the launch demo of the new version of ChatGPT earlier this month.
OpenAI responded with an official blog post on May 21, stating that the five voice options built into ChatGPT were provided by professional voice actors, and that the Sky voice option was not an intentional imitation of Scarlett Johansson, but that OpenAI was unable to share the identifying information of the voice actors for privacy purposes. Additionally, OpenAI said on X (formerly Twitter) that ChatGPT would temporarily disable the Sky voice option until further explanation.
It’s worth noting that OpenAI’s CEO had posted a dynamic on the day of the GPT-4o launch referencing Scarlett Johansson’s movie production of Her, in which the latter’s voice performance for the AI character was also a major reference in the current controversy.
Adobe Lightroom Goes Live with Generative AI Features
Adobe recently pushed Generative AI-based Lens Blur to Lightroom Stable users, and Object Removal, also based on Generative AI, is now available in Beta. Both features are powered by Adobe Firefly models and can be experienced in mobile, web and desktop applications.