Please explain the relationships and differences between, LaMDA, Bard, GPT-3, and chatGPT?


There is a lot of talk of large language models. However, I am a bit confused, especially when it comes to determining what is related to what and what are the similarities and differences. Is LaMDA more like like GPT-3 or chatGPT? Or am I failing to grasp something very fundamental?

In: 2

Think of it like this. Lamda and GPT-3 are backends, both built on googles Transformer architecture.

Bard and ChatGPT are products that use lightweight conversational models built on top of lamda and GPT-3, respectively. They both are models themselves, but also are the canonical names of the products.

It’s like Bitcoin to the blockchain. Bitcoin being a narrow use case of what blockchain could all be.