Bard is powered by a research large language model (LLM), specifically a lightweight and optimized version of LaMDA, and will be updated with newer, more capable models over time. It’s grounded in Google’s understanding of quality information. You can think of an LLM as a prediction engine. When given a prompt, it generates a response by selecting, one word at a time, from words that are likely to come next. Picking the most probable choice every time wouldn’t lead to very creative responses, so there’s some flexibility factored in. We continue to see that the more people use them, the better LLMs get at predicting what responses might be helpful.
While LLMs are an exciting technology, they’re not without their faults. For instance, because they learn from a wide range of information that reflects real-world biases and stereotypes, those sometimes show up in their outputs. And they can provide inaccurate, misleading or false information while presenting it confidently. For example, when asked to share a couple suggestions for easy indoor plants, Bard convincingly presented ideas…but it got some things wrong, like the scientific name for the ZZ plant.