ChatGPT 4 Starts Recognizing Human Faces, OpenAI Is Concerned
OpenAI has developed a powerful AI language model called ChatGPT 4, which can generate text, write essays, and now analyze images to provide detailed descriptions.
They allowed a blind man named Jonathan Mosen to test the visual analysis feature, and he found it very helpful in identifying items in his hotel room bathroom, such as shampoo, conditioner, and shower gel, along with their capacities and even the type of tiles in the shower.
However, OpenAI is now concerned about potential misuse of this feature, especially for facial recognition, which could raise privacy issues.
While the technology can identify public figures with Wikipedia pages, it is not as advanced as specialized facial recognition tools available on the internet.
The company worries that making this feature widely accessible could have legal implications in places where consent is required to use biometric information like facial features.
They are also cautious about the tool providing inappropriate assessments of people’s faces, which could lead to problems.
OpenAI aims to have open conversations with the public to understand their concerns and preferences. If people are uncomfortable with this feature, the company is willing to make adjustments accordingly.
Furthermore, there is a possibility that the visual analysis feature could generate false information, similar to how AI text models sometimes produce misleading results.
For example, if given a picture of someone on the verge of becoming famous, the model might mistakenly hallucinate a different name.
Also Read:- Hackers Trick With Fake Banking Apps, Exploits Android’s WebAPK Tech
The latest visual analysis feature of ChatGPT 4 is really impressive as well as shocking. It may assist visually impaired persons like Jonathan Mosen but it is also concerned about its potential misuse, especially for facial recognition.