OpenAI’s ChatGPT has been in the internet world since its release. Many people have set up their business with the help of this AI chatbot, many have even got good jobs by getting their resume from it. Now this chatbot is not only limited to text response, its new version i.e. GPT-4 has also started doing image analysis. GPT can give a description by looking at 4 pictures. Can answer questions related to it and identify people present in it. Users can take help in problems related to photographs.
ChatGPT is constantly trending, so it’s only natural that it will start doing image analysis along with text. For this he is getting enough training data from internet. However, OpenAI is cautious about the potential risks of image analysis. Since GPT-4 can recognize the faces of many famous people, the company wants to ensure that GPT-4 takes privacy and consent into account when analyzing and using images. Sandhini Agarwal, policy researcher at the company, says the company wants to be transparent with users about the use of its chatbot’s visual analysis capabilities. For this, the company is considering feedback from users, guidelines for usage and safety measures. The company is working towards creating rules for AI systems, so that people understand their responsibilities while using them.
Image analysis can prove helpful for the visually impaired. Its early users included Jonathan Mosen, who has been blind since birth. He is the CEO of Workbridge, an employment agency for people with disabilities. During a trip some time ago, he used GPT-4’s visual analysis feature.
Also Read: Should the fridge run continuously for 24 hours? Can you stop for 1-2 hours in between?
With this, he was able to distinguish between different dispensers in hotel bathrooms, which contained what and how much. It provides much better information than the usual visual analysis software. However, keeping consent and privacy in mind, the company has prevented the app from sharing information about people’s faces. Because of this, the app did not provide Mosen with information related to people’s faces.
Also Read: Girl Starts Undressing On Video Call, Boy Goes Brain Shut And Then…
One of OpenAI’s concerns regarding image analysis is that chatbots can give false information about an image or the person in it. It is possible that by posting a picture of a famous person at a particular event, GPT 4 can tell users about another famous person attending the same event. Such information may confuse users.
Violet & Daisy, a captivating action-comedy directed by Geoffrey Fletcher, revolves around the lives of…
MBC's latest release, the trailer for episode 5 of "Wonderful World," showcases the captivating performances…
Deadpool 3 & Wolverine Super Bowl Trailer Easter Eggs The Deadpool 3 Super Bowl trailer…
The Nagi Nagi no Mi is a Paramecia-type Devil Fruit with the unique ability to…
Recent images from the set of a canceled Game of Thrones spin-off have surfaced, showcasing…
The forthcoming installment in the Jurassic World movie series has been slated for release, along…