The main distinction between ChatGPT and Bard (now renamed Gemini) lies in their data sources. ChatGPT is trained on a fixed dataset that has not been updated since 2021. In contrast, Gemini is trained on an "infiniset" of data with access to the internet in real time, enabling it to retrieve current information.

ChatGPT is trained using a diverse range of internet text based on the data available as of September 2021. Therefore, ChatGPT’s responses are limited to information and events up until that time.  ChatGPT AI-generated content ranges from blog posts, summaries of transcripts and meetings, law briefs, to simple explanations of complex topics. 

Gemini.google.com has access to real-time information from the internet, however, that doesn’t mean all its information is 100% correct. Gemini’s own FAQs page states that “Bard (now renamed Gemini) is experimental, and some of the responses may be inaccurate, so double-check information in Bard's (now renamed Gemini) responses.”  Gemini’s goal is to retrieve the most current information and present it in a conversational type of response as opposed to a search engine response.

ChatGPT automatically saves all the chats on the sidebar for future reference; whereas Gemini loses all chats when closed, however it does have the ability to export chats prior to ending the session.

Both models are still in training and request feedback from users to build a better product/experience.