Large Language Models (LLMs) are revolutionizing how humans interact with computers, powering applications such as ChatGPT and DALL-E. These models understand language well, leveraging techniques like transformers, bidirectional encoding, and autoregressive models to generate content, find information, converse, or help organize data. LLMs have numerous use cases, including generative text, real-world applications in various industries, summarization, rewrite, search, question answering, clustering, classification, and more. Despite their potential, challenges such as hallucination, cost of creation and usage, interpretability, risk of spectacular fails, and impersonal results exist. The future of LLMs is promising, with trends like LLM overload, trust but verify, maturity curve, cheaper to build, and enterprise-ready features on the horizon. As these models become more powerful, they will change how people interact with computers, making almost every application powered by LLMs in a few years' time.