Perplexity AI has been ahead of its time by generating search results using a large language model like ChatGPT, but offering even better factuality, while connected with the internet. Now they have come up with an even better update. 

Ask any Twitter user, they will tell you how inconvenient the search feature is. Filtering out Tweets by specific people about one specific topic is nearly impossible. Perplexity AI is here to fix that problem of Twitter search with Bird SQL. 

BirdSQL is built on the large language model of OpenAI’s API, Twitter API, and PostgresSQL. The model translates natural language to SQL code through OpenAI’s codex. This makes it easier to search posts on Twitter. You can also search for specific Tweets of a specific person by sorting them out by likes and number of retweets. For example, you can search for, “Top 10 Tweets about Elon Musk in the last hour” and it will give you precise results. 

Click here to check Bird SQL.

What Bird SQL does might seem revolutionary, but this is what the Twitter community has been asking for a long time. Former CEO of Twitter, Jack Dorsey, also called the engine ‘great’.

ChatGPT with Internet

The section of the website, “Ask Anything”, as the name suggests, allows you to ask anything and it generates output while also citing the sources it retrieved the information from. It was inspired by OpenAI’s WebGPT.

In 2021, OpenAI created WebGPT, which was intended to improve the factual accuracy of large language models by connecting them to the web by fine-tuning GPT-3 for open-ended questions. This essentially gives a glimpse of what ChatGPT would look like if it was connected to the internet.