To summarize, the key differences between ChatGPT 4 and its predecessor, GPT 3.5, are:
Multimodal inputs: ChatGPT 4 can generate content from both image and text prompts, whereas GPT 3.5 takes only text prompts.
Longer responses: ChatGPT 4 can generate responses of more than 25,000 words, while GPT 3.5 is limited to about 3,000-word responses.
Style and tone flexibility: Developers can decide their AI's style of tone and verbosity in ChatGPT 4, whereas the previous iteration of the technology had a fixed tone and style.
Higher accuracy: ChatGPT 4 is 82% less likely to respond to requests for disallowed content than its predecessor and scores 40% higher on certain tests of factuality.
New capabilities: ChatGPT 4 has demonstrated the ability to help individuals calculate their taxes and create a website based on a hand-drawn mock-up. It also powers a virtual volunteer tool for the Be My Eyes app.
However, like its predecessors, ChatGPT 4 has limitations and can generate inaccurate responses known as "hallucinations." It also lacks knowledge of events that occurred after September 2021 and does not learn from experience. The image-input feature is not publicly available yet, and only ChatGPT Plus subscribers and software developers with a wait-list can access the text-input feature.