GPT-4
Why in News?
- Recently, OpenAI has launched its ChatGPT Plus subscription for Indian users, providing them with early access to the latest language model GPT-4.
Highlights
- According to OpenAI, GPT-4 is more advanced than its predecessors when it comes to creativity, visual comprehension and context.
- It also possesses the ability to collaborate with users on various creative projects, including music, screenplays, technical writing, etc.
- It can process up to 25,000 words of text and facilitate extended conversations.
- GPT-4 can encompass more than just text – it also accepts images as input.
- On the contrary, GPT-3 and GPT-3.5 only operated in one modality, text, allowing users only to ask questions by typing them out.
- GPT-4 is more multilingual and OpenAI has demonstrated that it outperforms GPT-3.5 and other Large Language Models (LLMs) by accurately answering thousands of multiple-choice across 26 languages.
- It handles English best with an 85.5% accuracy, but Indian languages like Telugu aren’t too far behind either, at 71.4%
- This move comes at a time when tech giants are competing to offer the best generative AI to the customers.
- ChatGPT is a variant of GPT (Generative Pre-trained Transformer) which is a large-scale neural network-based language model developed by OpenAI.
- GPT models are trained on vast amounts of text data to generate human-like text.
- It can generate responses to a wide range of topics, such as answering questions, providing explanations, and engaging in conversations.
- In addition to being able to “admit its mistakes, challenge false premises, and refuse unsuitable requests,” the ChatGPT can also “answer follow-up questions.”
- The chatbot was also trained using Reinforcement Learning from Human Feedback (RLHF).