GPT 4 is the latest release by OpenAI, the lab tech responsible for the popular text-to-image tool Dall-E and the even more popular natural language application ChatGPT. And it's an interesting one!
What makes GPT4 different is that it's multimodal AI, which can analyze both text and image prompts, to produce written-only results. But also, it's the lab's best-yet software regarding capability and stability.
Intrigued? Then read on for more info!
And if you want all the details on GPT4, who can access it, and how, read my dedicated article on Aisecrets.com!
What is GPT 4: AI that Interprets Language and Images
OpenAI’s latest AI model accepts prompts –user input or instructions– written or visual (such as photos, screenshots, diagrams, etc.) but produces text-only results.
Besides understanding written instruction, GPT 4 can identify and analyze an image's elements and utilize that interpretation to perform different tasks.
And it can do it with much more accuracy than ever before. According to OpenAI, this software has thrown the best-ever results during their tests. While they clarify it does not replace humans in real-world scenarios, they claim it reaches human-level performance results in different professional and academic environments.
What is Being Built with GPT 4: Apps that Assist Humans
The company focuses on the fact that this development isn't aimed at replacing humans in their jobs or their abilities but rather to help them, be it to improve workflows or assist them in areas where they need it.
For example, we learned that Microsoft's new Bing chatbot is using GPT 4 and that an assistance app for the visually impaired named Be My Eyes has developed a new Virtual Volunteer that can analyze images provided by users and answer questions or produce other relevant results from them –such as telling them what is inside their fridge and what they can cook with it.
Overall, it's a very interesting new technology and a new step into deep learning applied to everyday life.
