Normally, when we’re chatting with an AI chatbot, we do understand that we’re talking to a robot and not a human. This is because the answers are pretty formal and fixed. However, now that Google is introducing LaMDA, our conversations with chatbots might change too.
LaMDA stands for “Language Model for Dialogue Applications”. As the name suggests, it is a technology that can make conversations more natural with chatbots or voice assistants.
Back in 2017, Google had introduced Transformer, a machine learning model that they built for understanding languages. On this very model, they have recently built the LaMDA, that can understand language and make conversation on anything.
Interesting Features of Google LaMDA
One of the reasons why LaMDA has been generating excitement among so many people is the way it can make conversation on anything. Here’s what we can say about this technology:
- It has been trained to make open-ended conversations that can grow.
- Unlike other language and conversation based models, it can respond more like a person.
- It can literally talk on any subject.
- You may not be able to predict the response based on anything you ask.
- The technology will be used by all Google apps that make conversation with users in verbal or written form.
You can pull a conversation with an application that uses LaMDA with interest. Here’s an example that might explain why:
Suppose, you tell LaMDA about your achievement of coming first in class. While a normal robot wouldn’t say more than “Congratulations”, LaMDA might be able to respond this way:
“Wow, this is a great achievement, congratulations! How much did you get?”
This is the main difference between Google LaMDA and other technologies. It is capable of making a real conversation with its user.
