GPT-3 might sound like the name of an Android from Star Wars, but it’s really a neural network–powered language model — a computer program that, in layman’s terms, takes a load of text and then “speaks” by guessing what the next word is most likely to be.

OpenAI, the nonprofit behind the system, says it built GPT-3 as part of its goal to create a “friendly A.I.” that “benefits humanity” as a whole (as opposed to a Terminator-style unfriendly A.I. that… well, doesn’t.)

Generative Pretrained Transformer 3 (GPT-3) is an obscure-sounding mouthful that largely manifests itself as a sort of chatbot. You send it messages. It replies. And it’s actually good. Take, for example, this “conversation” between Kevin Lacker, CTO of Parse, and GPT-3:

Human:_ What is your favorite animal?_

GPT-3:_ My favorite animal is a dog._

Human:_ Why?_

GPT-3:_ Because dogs are loyal and friendly._

Human:_ What are two reasons that a dog might be in a bad mood?_

GPT-3:_ Two reasons that a dog might be in a bad mood are if it is hungry or if it is hot._

Its phrasing is a little stiff and slightly overly formal, as if it’s on best behavior. It’s like GPT-3 understands what it’s asked and figures out a reply. It seems like a person.

#machine-learning #tech #openai #ai #artificial-intelligence