For the past 20+ years, artificial intelligence and its early predecessor technologies have been helping organizations get things done. From machine learning algorithms that personalize customer experiences, to newer Generative AI models that write entire reports, the prospects for AI technology have grown more intriguing with each passing year.
Recently, the excitement of Large Language Models (LLMs) such as ChatGPT have unleashed a powerful new capability on the world. With a wide breadth of knowledge and the ability to perform multi-step reasoning, they can respond to a surprising variety of prompts. However, one problem remains. They have no agency. Generative LLMs respond only when prompted, rather than continuously ingesting new data, like humans do with our senses, or continuously working in service of a goal.
