Google has launched its AI Test Kitchen app to experiment with the LaMDA 2 conversational AI model announced at Google I/O this year. The MDA 2 (Language Model for Dialog Applications) is the latest version of the model released in 2021 following a lot of user feedback, and the AI Test Kitchen will continue to collect data from Google users as it continues to refine the model.
The idea behind LaMDA 2 is to create a comfortable AI model with more open conversations and unpredictable tangents in the way normal human conversations operate. The model grew out of Meena, a chatbot that Google rated as the most advanced when it was released in 2020. LaMDA 2 is trained in dialogue and is taught to detect relationships between words, sentences and longer passages so that it can better understand the context. of the conversation and reply with reference to the entire conversation, not just the last thing the user said.
AI Test Kitchen offers three different ways for users to play LaMDA 2. ‘Talk About It’ allows users to try to talk to the AI about a topic, and the AI will not be distracted by other topics. However, at this time, the focus is limited to dogs. The ‘List It’ tool uses LaMDA to break down any goal or objective into a series of tasks that will help achieve that goal, with the example of starting a garden by generating half a dozen suggestions for a list, including choosing a style, planting seeds , irrigation and pest control. The final test area is called ‘Imagine It’ and allows users to ask the AI to be creative in describing a concept shared by the user, a kind of short scene useful as a creative writing prompt.
“Like an actual test kitchen, the AI Test Kitchen will serve up a rotating set of experimental demos. These are not finished products, but are designed to give you an idea of what is being made possible with AI in a responsible way. Our first set of demos explores the capabilities of our latest version of LaMDA, which has undergone key security enhancements,” Google explained in a blog post. “As you try each demo, you’ll see LaMDA’s ability to generate creative responses on the fly. This is one of the strengths of the model, but it can also present challenges as some responses may be inaccurate or inappropriate. We have been testing LaMDA internally for the last year, which has produced significant quality improvements. More recently, we’ve done dedicated rounds of adversarial testing to find additional flaws in the model.”
Google Augments Helper Robot with natural language understanding
Google demonstrates a new conversational AI model and opens an AI test kitchen
Google upgrades the intuition and flexibility of conversational AI at Google I/O 2021 with LaMDA and MUM