News on Medial

AI doesn’t know much about golf. Or farming. Or mortgages. Or…

LivemintLivemint · 2m
AI doesn’t know much about golf. Or farming. Or mortgages. Or…

- Many AI models, including ChatGPT, lack deep, specific knowledge even on mainstream topics like golf. - Companies need to augment general models with industry-specific or business-specific data to make them useful. - Augmentation options vary in terms of accuracy, reliability, costs, and complexity. - Retrieval augmented generation (RAG) is an approach used by the PGA Tour to avoid AI errors. It includes organization-specific information in the model's input and involves human review of the model's outputs. - Fine-tuning parts of the model with proprietary data can result in higher accuracy and relevance of responses, but it can be more expensive and require specialized talent. - The highest level of accuracy is achieved when a company trains its own models, but this approach is cost- and talent-prohibitive for most enterprises.

Comments

Download the medial app to read full posts, comements and news.