OpenAI is introducing audio and image capabilities to enhance ChatGPT, expanding beyond written prompts. The update will be rolled out to paid versions of the application over the next two weeks, as announced in a recent blog post.
ChatGPT’s New Features: With this update, users will be able to engage in voice conversations with the chatbot, making it more akin to popular AI assistants like Siri and Alexa. This enables various applications, such as narrating bedtime stories, settling debates, and vocalizing text input.
OpenAI’s technology behind this update is also being utilized by Spotify to help podcasters translate their content into different languages.
Additionally, users can upload single or multiple images to the interface and employ a drawing tool to highlight specific image elements. This image feature facilitates troubleshooting, meal planning by examining fridge contents, or analyzing complex data graphs for work-related tasks.
Public Response: OpenAI’s announcement has generated mixed reactions on social media. While some users have celebrated the new features, others have voiced concerns.
Trevor Darrell, a professor at UC Berkeley and a co-founder of Prompt AI, mentioned the “uncanny valley gap” — the fear of AI becoming too human-like. Complex interfaces that don’t perfectly mimic human interaction might feel strange to some users, potentially affecting usability.
Moreover, some users have expressed concerns about OpenAI’s recent legal issues, warning against using ChatGPT due to copyright infringement and intellectual property rights violations.
There are also concerns that these updates could potentially replace smaller AI startups, software engineers, and educators in the future.
The introduction of AI-generated voices raises fears of deepfakes, voice scams, and identity theft. AI voice scams are becoming more prevalent, leading to financial losses for victims.
The addition of voice recognition may make the feature less accessible for people who speak with non-mainstream accents.
Concerns also arise with the image function potentially enabling the AI to bypass image verification CAPTCHA tests on websites, which are designed to limit access.
A recent study, yet to be peer-reviewed, suggests that AI bots can solve CAPTCHA tests more quickly and accurately than humans.
OpenAI’s Response to Risks: OpenAI acknowledges the potential for malicious actors to exploit the voice feature for fraud and impersonation. To mitigate this, the company specifies that the technology is designed for specific use cases, such as voice chats created with voice actors they have directly worked with.
OpenAI also acknowledges the limitations associated with images in AI, including the generation of false information about images. They have implemented technical measures to limit ChatGPT’s ability to analyze and make direct statements about individuals in images.
Check out the latest news in our Global News section
Stay updated on environmental data and insights by following KI Data on Twitter