Meta has announced the launch of its AI research project ‘BlenderBot 3’, a chatbot which the company says can search the internet to talk about “nearly any topic”.
The technology giant said that it will share the model, data, and code for the AI chatbot with scientists to help advance conversational AI.
The Facebook-owner explained that the technology is designed to improve its conversational skills and safety through feedback from people who chat with it, focusing on helpful feedback while "avoiding learning from unhelpful or dangerous responses".
The company explains that because AI chatbots are known to sometimes mimic and generate unsafe, biased or offensive remarks, it has conducted large-scale studies, co-organised workshops and developed new techniques to create safeguards for BlenderBot 3.
The current model can still make rude or offensive comments, which is why the company says it is collecting feedback to improve on future chatbots.
It says that its BlenderBot series has made progress in conversational skills like personality, empathy, and knowledge - incorporating long-term memory, and searching the internet to carry out meaningful conversations.
Recent Stories