Google AI blog:
While conversations tend to revolve around specific topics, their open-ended nature means they can start in one place and end up somewhere completely different. A chat with a friend about a TV show could evolve into a discussion about the country where the show was filmed before settling on a debate about that country’s best regional cuisine.
That meandering quality can quickly stump modern conversational agents (commonly known as chatbots), which tend to follow narrow, pre-defined paths. But LaMDA — short for “Language Model for Dialogue Applications” — can engage in a free-flowing way about a seemingly endless number of topics, an ability we think could unlock more natural ways of interacting with technology and entirely new categories of helpful applications.
LaMDA was shown off in the Google I/O keynote, embedded below (it’s the second video). The discussion starts at about 17:03 in. Obviously, this is a demo, and not a shipping product. It’s interesting to me how free Google is in showing off internal technologies, contrasted with Apple’s focus on products that are for sale or coming soon. Of course, there are exceptions on both sides, but the trend is clear.
LaMDA is a tantalizing demo, showing off a technology that is so very close to Apple’s original Knowledge Navigator personal assistant concept, first shown off back in 1987. If you’ve not seen it, check it out below (it’s the first video).
The differences between Knowledge Navigator, the LaMDA demo, and Siri are stark. Of course, Siri is a shipping product, and has both limited context (once you move on, the stuff you asked about is forgotten) and a limited domain set (Siri knows about a limited set of things. If Siri doesn’t have the topic built-in, it defaults to a web search, or an “I can’t help you” response).
LaMDA, on the other hand, is a highly controlled demo, used internally, an R&D project. Again, not a shipping product.
But, that said, both Knowledge Navigator and LaMDA really make me want much more from Siri. I want more sophisticated language, more context, more understanding. Something more human, more beyond-the-moment interaction. I’d like a little more conversation.
In this one instance, I’d love to see what experiments the Siri team is working on, get a little taste of where Siri is headed.