Skip to main content

AI Pair Programming Showdown: ChatGPT vs Google Bard vs Bing Chat

· 2 min read
Morgan Moneywise
Morgan Moneywise
CEO at Morgan Moneywise, Inc.

Ever since large language models have taken the world by storm, I've been curious about how helpful these AI-powered chatbots can be to programmers like myself.

To put them to the test, I decided to build a couple of simple API endpoints using FastAPI and see how ChatGPT, Google Bard, and Bing Chat can assist me in the process.

Here's how they fared:

There are a few things that stood out to me from this experiment:

  1. ChatGPT, while highly capable, is limited by its knowledge cutoff date (September 2021). This means it can't help you with libraries or features introduced after that point.

  2. Bing Chat, on the other hand, has search engine features that make it more adept at assisting with newer libraries or features.

  3. Interestingly, Bing Chat can be convinced that its search index is not up-to-date and be persuaded to follow our command.

  4. Google Bard stands out for its speed in rendering output but tends to have the most errors among the three.

  5. One common issue with all three tools is their susceptibility to confidently hallucinating answers, sometimes even making up sources.

If you have any questions or comments, feel free to drop a line!