If you are charmed by the ChatGPT for its human-like answers, you are in for a bigger surprise. OpenAI, the makers of ChatGPT, said they have launched GPT-4, which is more intelligent, bringing in more reasoning and context to the answers that it gives.
GPT-4 is available on ChatGPT Plus (a paid version of ChatGPT). It also made APIs available, letting third parties to plug it into their products.
The firm has factored in the ChatGPT users’ feedback to improve GPT-4’s behavior. It also roped in over 50 experts for early feedback in domains, including AI safety and security.
Visual element and other featuers
In the latest version, OpenAI brings in the visual element, allowing the usage of visuals in your queries. The algorithm can intelligently read the image, understand the context and churn out an answer in no time.
OpenAI gives a peek view of the upcoming AI-based chat solution in its latest blog.. ”You can ask it to explain the plot of Cinderella in a single sentence where each word has to begin with the next letter in the alphabet from A to Z, without repeating any letters,” it says.
The artificial intelligence solutions firm claims GPT-4 can solve difficult problems with greater accuracy, thanks to its broader general knowledge and problem-solving abilities.
The firm said it spent six months to make GPT-4 safer and more aligned. “It is 82 per cent less likely to respond to requests for disallowed content and 40 per cent more likely to produce factual responses than GPT-3.5 on our internal evaluations,” it observed.
“GPT-4 is more creative and collaborative than ever before. It can generate, edit, and iterate with users on creative and technical writing tasks, such as composing songs, writing screenplays, or learning a user’s writing style,” it says.
It says GPT-4 surpasses ChatGPT in its advanced reasoning capabilities. “The difference comes out when the complexity of the task reaches a sufficient threshold—GPT-4 is more reliable, creative, and able to handle much more nuanced instructions than GPT-3.5,” it says.
OpenAI said it tested on a variety of benchmarks, including simulating exams that were originally designed for humans. It used publicly-available tests such as Olympiads to understand the difference between the two models.
OpenAi said it has worked with organisations like Morgan Stanley, Khan Academy, Government of Iceland and Duolingo to test the solution on their respective products.