What is the difference between Bard and ChatGPT?

OpenAI’s ChatGPTChat Generative Pre-Trained Transformer became an internet sensation when it was launched in November 2022. Since then, Google has introduced its AI-powered chatbot, Bard. These two are the most prominent AI chatbots available in 2023. Both companies, OpenAI and Google, admit that these AI tools aren’t as perfect as they should be, but an existing future of AI bots is clear because of them. While both of these are language models, there are key differences in the two. Some of them are listed below:

Input and output

Bard’s input and output are based on dialogues, while ChatGPT’s input and output are text-based. Bard uses human language to express thoughts, feelings, and stories, while ChatGPT uses natural language processing to interpret and generate text.

Training

ChatGPT uses the GPT-3 language model, while Bard uses LaMDALanguage Model for Dialogue Applications. GPTs are trained on tons of data using unsupervised pre-training and fine-tuning. LaMDA, on the other hand, was trained on dialogue, not text. 

Data source

Another difference between the chatbots is the data source. Bard is trained on LaMDA, which is why it can continually draw information from the internet in real-time, while ChatGPT’s sources are based on data from 2021. This can be seen in the image below, where the answer generated from Bard is on the right side, and the left side contains the answer from ChatGPT.