Recently, Google released its new AI chatbot named LaMDA 2 after the release of its AI Test Kitchen. The bot will chat with people on multiple subjects. Google has opened the registrations for all the potential people who would like to use this AI chatbot. It will be done via an app that the users can download to have full access. But right now Google is only letting people in the US sign up for this great opportunity.
On the other hand, Meta also made a similar chatbot named Blenderbot 3 that the public could use but it was not much of a success because people could make this AI bot say anything creepy and untruthful and other people were starting to believe it. The bot was even making lies about its founder, Mark Zuckerberg. Now when Google will start testing its AI bot with the public, Google will get to know the public reaction and how much this chatbox will be able to serve its purpose.
Many tech companies do not like to test their chatbox with the public because they think that it can severely damage the company's image if it doesn't work properly. But if they want to improve their bot's working, it is best to open them to public use so the bots get an open environment to chat with people who can double trick them. This is solely for the betterment of the AI as the engineers who develop them couldn't get to the depth of it, even if they try as hard as possible.
Mary Williamson who is a research engineering manager at Meta says that if the bots will say unhelpful things, they are of no use to any of us. And by releasing AI chatbots to the public, we are pushing our way forward with more research.
There are some huge differences between the chatbots of Meta and Google as Meta released its AI bot freehand but Google is trying out some restrictions to make it work properly.
LaMDA will have three modes:Imagine It, Talk About it and List It and each mode will have its separate purpose. It's safe to say that this AI chatbot will have some limitations to it that will help to not spread false accusations.
Read next: Here Are the Top Google Search Trends for 2022
On the other hand, Meta also made a similar chatbot named Blenderbot 3 that the public could use but it was not much of a success because people could make this AI bot say anything creepy and untruthful and other people were starting to believe it. The bot was even making lies about its founder, Mark Zuckerberg. Now when Google will start testing its AI bot with the public, Google will get to know the public reaction and how much this chatbox will be able to serve its purpose.
Many tech companies do not like to test their chatbox with the public because they think that it can severely damage the company's image if it doesn't work properly. But if they want to improve their bot's working, it is best to open them to public use so the bots get an open environment to chat with people who can double trick them. This is solely for the betterment of the AI as the engineers who develop them couldn't get to the depth of it, even if they try as hard as possible.
Mary Williamson who is a research engineering manager at Meta says that if the bots will say unhelpful things, they are of no use to any of us. And by releasing AI chatbots to the public, we are pushing our way forward with more research.
There are some huge differences between the chatbots of Meta and Google as Meta released its AI bot freehand but Google is trying out some restrictions to make it work properly.
LaMDA will have three modes:Imagine It, Talk About it and List It and each mode will have its separate purpose. It's safe to say that this AI chatbot will have some limitations to it that will help to not spread false accusations.
Read next: Here Are the Top Google Search Trends for 2022
