OpenAI’s ChatGPT has been concealing a secret this whole time: it possesses the capability to perceive and identify faces, reports The New York Times. At least, this has been the case since the release of GPT-4. This isn’t just another update folks; it’s a game-changing twist to our AI interaction story. Our beloved word wizard has essentially evolved to understand not only what we’re saying but how we’re saying it – smirk, raised eyebrow, and all. This new feature allows the model to interpret facial expressions and respond accordingly, making the interaction more personalized and human-like. For instance, if the user looks confused, the AI can detect this and offer to explain the topic in a simpler way. This could be a significant step forward in making AI more accessible and user-friendly. Does anyone feel like OpenAI is throwing us into a real-life sci-fi plot? Image: Cisco We’re talking about privacy and data security here, people. Our faces aren’t just our meal ticket anymore. Nope, they’re currency in this bold new world of facial recognition technology. And guess what? It’s shaky ground we’re stepping on, with a ton of uncertainty about how it can be exploited in the wrong hands. When I demonstrated the new AI-can-see-images feature Microsoft is rolling out for GPT-4/Bing, everyone suggests the same challenge: "give it a picture of what is in your fridge."I gave it a hard version, once normal, once asking for creative solutions. I think it pitied me. pic.twitter.com/U6Q860mhDH— Ethan Mollick (@emollick) June…ChatGPT can recognize faces – OpenAI just isn’t promoting it…yet