#ai #llm Created at 2023-01-27 # [Anonymous feedback](https://www.admonymous.co/louis030195) # [[Epistemic status]] #shower-thought Last modified date: 2023-01-27 Commit: 0 # Related - [[Computing/Intelligence/Machine Learning/Embedding is the dark matter of intelligence]] - [[Computing/Intelligence/Joint embedding]] - [[Computing/Human embeddings]] - [[Computing/Ideas/Multimodal AI assisted knowledge management]] - [[Computing/Intelligence/Machine Learning/Language model]] - [[Computing/Intelligence/Machine Learning/BERT]] # Embeddings In 2023, software without embeddings is like a smartphone without internet. Embeddings is the new way to query data, information generation increase nonlinearly and there is this unmet need to provide tools to work with this massive unstructured data easily Embeddings are an intermediary language between man and machine Embeddings is the bridge between man model's of the world and machine model's of the world Embeddings is like the line on a map, where the territory is our world Embedding is a bicycle for the mind Embedding is the bridge between [[Software 2.0]] and [[Software 3.0]] Embedding is to software what [[The Internet]] is to computers ## ELI5 Embeddings in machine learning are a way to represent words or phrases in a way that a computer can understand, like a mathematical formula. This makes it easier for the machine to recognize patterns in language. For example, if a machine is given a sentence like "The cat sat on the mat", it can understand that the words "cat" and "mat" are related, even if it doesn't know the exact meaning of the words. Embeddings help the machine to understand the context of a sentence and can be used to create better language models. https://twitter.com/CohereAI/status/1618237043116593152