Meta develops conversational AI chatbot: Concern for data-sharing?

Meta is calling this chatbot BlenderBot3 and made this chatbot available only for the people residing in the US at the moment.

According to Meta, this chatbot can engage in a general conversation or even help people with responses to queries people ask any digital assistant available in the market.

Meta’s BlenderBot 3 is a prototype, yet. It is built on the large language models that Meta has previously worked on. It is believed that LLMS are said to be a strong but flawed text-generation software. This AI conversational bot has been trained on a wide array of datasets of texts and can be utilised to generate a language.

Meta has announced that it wants people to use this bot and check its capabilities. The BlenderBot 3 can search the internet and provide specific answers, and even reveal its source of information. The technology giant claims that it wants to iron out the issues of the LLMS by making the AI chatbot talk to the general public in order to reduce the chances of BlenderBot 3 using unparliamentary language, slurs or making any insensitive comments.

People interested in trying Meta’s AI-based chatbot would have to give their consent for their data to be collected, according to Meta. The technology giant says that it will store this data and later publish it to the general AI research community.

Previous similar sort of experiments by Meta using chatbots did not turn out as per the expectations. Therefore, the technology giant wants to create better AI chatbots and generate productive outcomes.

It is noteworthy that Meta has always been into controversies regarding the data privacy concerns. It is totally dependent upon the people to give their consent to Meta for interacting with Blender Bot 3. As claimed by Meta, collected by the social media platform would further be sent for research purposes to the AI research community.

First article