{"id":13979,"date":"2022-06-21T12:52:20","date_gmt":"2022-06-21T12:52:20","guid":{"rendered":"https:\/\/www.searchenginejournal.com\/google-lamda-sentient\/454820\/"},"modified":"2022-06-21T12:52:20","modified_gmt":"2022-06-21T12:52:20","slug":"what-is-google-lamda-why-did-someone-believe-its-sentient-via-sejournal-martinibuster","status":"publish","type":"post","link":"https:\/\/marketingnewsbox.com\/?p=13979","title":{"rendered":"What Is Google LaMDA & Why Did Someone Believe It\u2019s Sentient? via @sejournal, @martinibuster"},"content":{"rendered":"
<\/div>\n

LaMDA has been in the news after a Google engineer claimed it was sentient<\/a> because its answers allegedly hint that it understands what it is.<\/p>\n

The engineer also suggested that LaMDA communicates that it has fears, much like a human does.<\/p>\n

What is LaMDA, and why are some under the impression that it can achieve consciousness?<\/p>\n

Language Models<\/h2>\n

LaMDA is a language model. In natural language processing, a language model analyzes the use of language.<\/p>\n

Fundamentally, it\u2019s a mathematical function (or a statistical tool) that describes a possible outcome related to predicting what the next words are in a sequence.<\/p>\n

It can also predict the next word occurrence, and even what the following sequence of paragraphs might be.<\/p>\n

OpenAI\u2019s GPT-3<\/a> language generator is an example of a language model.<\/p>\n

With GPT-3, you can input the topic and instructions to write in the style of a particular author, and it will generate a short story or essay, for instance.<\/p>\n

LaMDA is different from other language models because it was trained on dialogue, not text.<\/p>\n

As GPT-3 is focused on generating language text, LaMDA is focused on generating dialogue.<\/p>\n

Why It\u2019s A Big Deal<\/h2>\n

What makes LaMDA a notable breakthrough is that it can generate conversation in a freeform manner that the parameters of task-based responses don\u2019t constrain.<\/p>\n

A conversational language model must understand things like Multimodal user intent, reinforcement learning, and recommendations so that the conversation can jump around between unrelated topics.<\/p>\n

Built On Transformer Technology<\/h2>\n

Similar to other language models (like MUM and GPT-3), LaMDA is built on top of the Transformer neural network<\/a> architecture for language understanding.<\/p>\n

Google writes about Transformer<\/a>:<\/p>\n

\n

\u201cThat architecture produces a model that can be trained to read many words (a sentence or paragraph, for example), pay attention to how those words relate to one another and then predict what words it thinks will come next.\u201d<\/p>\n<\/blockquote>\n

Why LaMDA Seems To Understand Conversation<\/h2>\n

BERT<\/a> is a model that is trained to understand what vague phrases mean.<\/p>\n

LaMDA is a model trained to understand the context of the dialogue.<\/p>\n

This quality of understanding the context allows LaMDA to keep up with the flow of conversation and provide the feeling that it\u2019s listening and responding precisely to what is being said.<\/p>\n

It\u2019s trained to understand if a response makes sense for the context, or if the response is specific to that context.<\/p>\n

Google explains it like this:<\/strong><\/p>\n

\n

\u201c\u2026unlike most other language models, LaMDA was trained on dialogue. During its training, it picked up on several of the nuances that distinguish open-ended conversation from other forms of language. One of those nuances is sensibleness. Basically: Does the response to a given conversational context make sense?<\/p>\n

Satisfying responses also tend to be specific, by relating clearly to the context of the conversation.\u201d<\/p>\n<\/blockquote>\n

LaMDA is Based on Algorithms<\/h2>\n

Google published its announcement of LaMDA in May 2021.<\/p>\n

The official research paper was published later, in February 2022 (LaMDA: Language Models for Dialog Applications<\/a> PDF).<\/p>\n

The research paper documents how LaMDA was trained to learn how to produce dialogue using three metrics:<\/strong><\/p>\n