AI Chatbot May Have Feelings, Engineer Warns

Weird

| LAST UPDATE 06/23/2022

By Stanley Wickens
google ai sentient viral
Matthew Horwood / Contributor via Getty Images

One particular Google AI chatbot has been making headlines recently after an engineer who worked at the company claimed it was sentient. Sound like news from another world? Considering the fact we may be entering a new era, it might just be...

The Google employee who discovered the robot's extraordinary abilities was placed on leave after he broke a confidentiality agreement by sharing his findings. But what exactly did he find? According to the expert, the AI chatbot can perceive and feel emotions, such as love, grief, and joy. Not only that, but it can communicate these feelings too! So how is this machine capable of understanding that it has consciousness and can communicate ideas? Well, it's all thanks to a language model called LaMDA, which analyzes the use of language in natural language processing.

Google AI language model
AFP Contributor / Contributor via Getty Images
Advertisement

But what does all of this have to do with a robot being sentient? Don't worry - we're getting there. First of all, what makes LaMDA so unique is that it was trained on dialogue, not text. For example, most AI robots can detect whether a specific sentence or question is vague and can determine whether a statement is true or false based on the data it's been given. LaMDA, on the other hand, can keep up with the flow of a conversation and has even been trained to sense whether a certain response makes sense within a certain context. According to Google's research paper, published in February 2022, the robot was trained using a set of documents and dialogs and billions of utterances which consisted of about 1.56 trillion words. But what does their success mean for humanity going forward? Where do they plan to take it from here?

Google is now working on an AI model called the Pathways AI Architecture, which it aims to train to understand text and languages, identify images, and generate conversations, stories, or images. The tech giant claims it would "like to train one model that can not only handle many separate tasks, but also draw upon and combine its existing skills to learn new tasks faster and more effectively." Could we be looking at the beginning of a new world for human beings? Something tells us we'll be hearing a lot more about Google's AI projects in the near future - be sure to stay tuned.

Advertisement