OpenAI may release its latest large language model (LLM), GPT-4 (Generative Pre-Trained Transformer 4), as early as next week. OpenAI, the creator of well-known language models like ChatGPT and Dall-E, has introduced new features in GPT-4 such as multimodality, video processing, and the ability to generate AI-based videos from simple text prompts. The Microsoft-supported startup aims to expand the potential of language models with these cutting-edge advancements. At present, only text-based responses are supported by ChatGPT and other technologies that rely on GPT-3.5.
In the last two years, Open AI has completely revamped its deep learning infrastructure and collaborated with Azure to develop a supercomputer specifically designed for its needs. The GPT-4 training run was remarkably stable.
What can GPT–4 do?
As compared to the existing large language models and many state-of-the-art models that require benchmark-specific tailoring or extra training procedures, GPT-4 exhibits significantly superior performance. Furthermore, it is anticipated to have several cutting-edge capabilities that distinguish it from its predecessors, including:
Multimodality: GPT-4 will be capable of sourcing information from various sources, such as images, videos, and music, in addition to text.
Faster response times: It is anticipated that GPT-4 will address the sluggish response times that some users have reported experiencing with previous LLMs.
Improved human-like responses: GPT-4 is expected to generate more human-like responses to user queries, which could result in a more seamless user experience.
Higher capacity: GPT-4 is expected to have a greater capacity for processing and retaining information, enabling it to perform more complex tasks and provide more accurate responses.
Steerability: In contrast to the classic ChatGPT personality that has a fixed verbosity, tone, and style, while using GPT-4, developers can prescribe their AI’s style and task by describing those directions in the “system” message.
GPT-4 is anticipated to be a significant improvement over its predecessors, with advanced capabilities that could revolutionize natural language processing and improve the efficiency of various applications, such as chatbots and language translators.
OpenAI is reportedly developing a mobile app powered by GPT-4. It is worth noting that ChatGPT, a web-based language model, currently does not have a mobile app. Microsoft and OpenAI have not revealed any plans to integrate GPT-4 into Bing search, however, it will likely be utilized in Bing chat. While ChatGPT is based on GPT-3.5, Microsoft’s Bing search leverages a combination of GPT-3, GPT-3.5, and their proprietary technology, Prometheus, to generate answers quickly while utilizing real-time information.
The growing popularity of ChatGPT
ChatGPT has been well-received by users, and many have found it to be a useful tool for generating human-like responses to their queries. It had attracted millions of users in just five days. The platform uses the powerful GPT-3.5 technology, which has been trained on massive data sets and is capable of responding in various styles and languages, making it one of the most advanced language processing models available.
What makes this AI chat tool quite popular is its ability to respond in natural-sounding language, making it indistinguishable from a conversation with a real human. Users have been amazed at its ability to perform tasks beyond basic conversations, such as writing articles, academic papers, and even job applications, as well as assisting with coding.
Subscribers of ChatGPT Plus will have access to GPT-4 on chat.openai.com, but with a usage limit. The precise usage cap will be adjusted by OpenAI based on demand and system performance, but it will initially be restricted due to capacity constraints.
OpenAI may introduce a new subscription level for those who require higher-volume GPT-4 usage, depending on traffic patterns. Additionally, it may offer a limited number of free GPT-4 queries to allow non-subscribers to try it out.
OpenAI continues to work on improving ChatGPT and other language models, incorporating new features and advancements to enhance their functionality and address any issues users may be experiencing.
Read next: Microsoft empowers network operators by introducing Azure Operator Nexus
