Meta has developed an AI chatbot and made it available to the public on the internet. This chatbot has been created by Meta’s AI research labs, which wants to gain feedback about the chatbot from the public.
This chatbot from Meta is called BlenderBot 3, which is hard to be confused with anything but a chatbot. And Meta has made this chatbot available only to the people residing in the US for now.
The company has stated that the BlenderBot 3 can engage in a general conversation, or even help you with responses to queries that you would ask any digital assistant available in the market.
BlenderBot 3 has been developed as a prototype and built on the large language models that Meta has previously worked on. LLMS are said to be a powerful but flawed text-generation of software. BlenderBot 3 has been trained on a wide array of datasets of texts which can be utilised to generate a language.
Meta wants to use people to test the capabilities of the BlenderBot 3, which can search the internet and give you specific answers, and even tell you its source of the information. Meta wants to iron out the issues of the LLMS by making the AI chatbot talk to regular people in order to minimise the chances of BlenderBot 3 using vulgar language, slurs or making any insensitive comments.
Those keen on trying out the BlenderBot 3 chatbot will have to give consent for their data to be collected, which Meta says, will be stored and later published to be used by the general AI research community.
Previous experiments with chatbots made available to the public did not deliver the expected results. So, it would be intriguing to see how Meta learns from past mistakes and uses the BlenderBot 3 to attain a satisfying outcome.