It is highly probable that OpenAI, the developer of the ChatGPT chatbot, will unveil GPT-4 next week, which is the latest iteration of their large language model. GPT-4, supported by Microsoft, will introduce novel capabilities, such as multimodality, video processing, and the capability to generate AI-generated videos from basic text prompts. Presently, ChatGPT and other technologies powered by GPT-3.5 can solely produce text-based replies.
According to German news website Heise, it will be more powerful than GPT-3.5, which currently powers ChatGPT. The report said GPT-4 will include multimodality, an ability to function in multiple modes such as text, images, and sounds.
In an interview to Heise, Microsoft Germany’s Chief Technology Officer (CTO) Andreas Braun stated, “We will introduce GPT-4 next week … we will have multimodal models that will offer completely different possibilities — for example videos.”
Aside from its multimodal capabilities, GPT-4 may be successful in resolving ChatGPT’s problem of slow response to user-generated queries.
OpenAI has previously launched a tool named ‘ZeroGPT’, which enables users to identify if a given text was generated by an artificial intelligence (AI) program and presented as if it was written by a human, or if it was truly authored by a human being.
ZeroGPT uses DeepaAnalyze technology to determine the source of the input. Currently, its accuracy rate exceeds 98%, but OpenAI is working to reduce the error rate to less than 1%.
Users must enter the text into the box and then click ‘Detect Text,’ according to OpenAI. The AI software will check the input and publish the result using algorithms developed by the ZeroGPT team.
ChatGPT is a chatbot that uses artificial neural networks to answer questions in a human-like manner. The viral chatbot was released in November 2022 by San Francisco-based OpenAI had recently passed the milestone of 100 million active monthly users only two months after its launch.
