OpenAI CEO Sam Altman spoke at an event held in Seoul, South Korea on Friday, June 9, 2023.
Bloomberg | Bloomberg | Getty Images
OpenAI’s ChatGPT can now “see, hear, and talk.” Or at least it can understand spoken words, respond with synthesized voices, and process images, the company says. announced Monday.
The chatbot update – the biggest for OpenAI since the introduction of GPT-4 – allows users to select voice conversations in ChatGPT’s mobile app and choose from five different synthesized voices for the bot to respond to. Users can also share images on ChatGPT and highlight areas of focus or analysis (think: “What types of clouds are these?”).
OpenAI says the change will roll out to paid users within the next two weeks. Audio functionality is limited to iOS and Android apps, but image processing functionality is available on all platforms.
Driving this major capability is OpenAI, microsoft, Google And humanity. In an effort to encourage consumers to incorporate generative AI into their daily lives, major technology companies are racing to release new chatbot apps as well as new features, especially this summer. Google announced a number of updates to its Bard chatbot, and Microsoft added visual search to Bing.
Earlier this year, Microsoft increased its investment in OpenAI, adding an additional $10 billion, making it the year’s largest AI investment, according to PitchBook. The startup reportedly ceased operations in April. $300 million stock sale It is valued at $27 billion to $29 billion, and investors include Sequoia Capital and Andreessen Horowitz.
Experts have expressed concerns about AI-generated synthetic voices. In this case, users not only have a more natural experience, but they may also be able to do more. Convincing deepfakes. cyber attacker Researchers have already begun investigating how deepfakes can be used to penetrate cybersecurity systems.
OpenAI acknowledged those concerns in a statement Monday, saying the synthesized voices were “created using voice actors we worked with directly” and not collected from strangers.
The release also provided little information about how OpenAI would use consumers’ voice input or how the company would protect that data if it were used. OpenAI did not immediately respond to a request for comment, but its terms of service state that consumers own their input “to the extent permitted by applicable law.”