What is OpenAI's GPT-4 model?

GPT-4 is "multimodal," meaning it can answer questions that include both images and text.

Photo: GI

GPT-3.5 only provides results from written queries, whereas the latest version of this large language model can use images as input data to identify and analyze objects in images.

GPT-3.5 is limited to approximately 3,000 words in responses, while GPT-4 allows responses of over 25,000 words.

GPT-4 is 82% more responsive to requests for inappropriate content than its predecessor and scores 40% higher in several practical tests.

It will also allow developers to decide on the tone style and level of detail for their AI.

OpenAI says ChatGPT users will soon have the option to change the chatbot's tone and response style.

The latest version has outperformed its predecessor in both the U.S. bar exam and the Graduate Resume Exam (GRE).

According to OpenAI, GPT-4 has similar limitations to its predecessors and is "inferior to humans in many real -world situations."

Inaccurate responses, known as "hallucinations," pose a challenge to many AI programs, including GPT-4.

GPT-4 lacks knowledge of events that occurred after September 2021.

GPT-4 is being integrated into Microsoft's Bing AI chatbot and the popular language learning platform Duolingo.

Mai Anh (according to Reuters)

Source: https://www.congluan.vn/mo-hinh-tri-tue-nhan-tao-gpt-4-cua-openai-la-gi-post239471.html