Last September we wrote a blog about the race to develop large language models and the recent advances in natural language processing. Two months later, in November, OpenAI allowed public access to GPT-3, its state-of-the-art NLP. Now that this powerful technology is available for anyone to try for free, we’d like to offer a quick note on what you should know about GPT-3 and its potential uses.

“GPT-3 is a text-generation algorithm that is capable of producing human-like text.

GPT-3 is based on a powerful machine learning technique known as a transformer architecture. This architecture has allowed the model to learn from large amounts of unlabeled text data, such as books, articles, and social media posts. This has enabled the model to generate text with a higher level of accuracy and complexity than earlier models.

GPT-3 can be used to create content such as articles, essays, and blog posts. It can also be used to generate descriptions for images and videos, create chatbots that are able to generate contextual conversations, and generate personalized product recommendations. Additionally, GPT-3 can be used to create machine translation applications that can translate text from one language to another with high accuracy. GPT-3 can also be used for natural language processing tasks such as text summarization, classification of text, question-answering, and sentiment analysis. Finally, it can be used to create natural language understanding applications such as semantic search, text analysis, and part-of-speech tagging.

Overall, GPT-3 is an impressive and powerful NLP model that is capable of producing human-like text with impressive accuracy and complexity. It has already been used in many applications, and its potential for creativity and innovation is immense.”

Now for the reveal. This blog post is an example of what you can do with NLP’s like GPT-3. The text description of GPT-3 in italics above was actually generated by an NLP in response to our request to “list the uses of GPT-3.”  It took the NLP, Chatsonic, less than 3 seconds to come up with the answer.

We decided to move beyond a simple informational question to test the machine’s “creativity.” We asked it to write a poem about GPT-3 in the style of E.E. Cummings. A few seconds later we had the composition below:

GPT-3 so vast, so mighty and tall

The most impressive AI of them all

Doing tasks with ease and speed

It’s the ultimate in AI technology indeed

So much to learn, so much to explore

GPT3 is the future, that much is for sure

From businesses to academics and beyond

GPT3 will be leading the charge for the song

It is the harbinger of a new dawn

Where algorithms will be relied upon

GPT3 will be the beacon of light

That will guide us into a brand new night


Not bad for a machine.