Small model approach could be more effective than LLMs

Small language models are more reliable and secure than their large counterparts, primarily because they draw information from a circumscribed dataset. Expect to see more chatbots running on these slimmed-down alternatives in the coming months.