OpenAI: GPT-3 Language Model Generates 4.5 Billion Words Daily

The GPT-3 language model, created by the non-profit company OpenAI, generates 4.5 billion words every day. This was reported by the developers.

In the 9 months since we launched GPT-3, developers have created extraordinary apps using our API. Today GPT-3’s search, conversation, and text completion powers 300 applications focused on new ideas from productivity and education to creativity and games. https://t.co/UGk7yVb1H9– OpenAI (@OpenAI) 

March 25, 2021

OpenAI also noted that the model achieved such results thanks to more than 300 applications and tens of thousands of developers. 

For example, startup Viable uses GPT-3 to analyze customer reviews, identify emotions and sentiments from surveys, support calls, or reviews. 

Fable Studio is using the OpenAI language model to create virtual reality dialogues, and Algolia is using it to improve search products.

According to some analysts, it is a bad idea to base a startup on someone else’s technology: the ease of using GPT-3 to create a product will increase the number of competitors. At the same time, no company will benefit from this technology as much as OpenAI itself, skeptics are sure. 

Experts also see a benefit for Microsoft, which in September 2020 received the exclusive right to use the language model on the Azure cloud platform.

Recall that in January 2021, OpenAI showed a DALL-E algorithm based on GPT-3, which generates realistic images of non-existent things based on a text description.

Google previously unveiled its 1 trillion parameter language model, six times the GPT-3.

Leave a comment