Meta Releases Its New AI Chatbot BlenderBot 3
Meta has just announced the release of its latest AI Chatbot, BlenderBot 3, a conversational agent that is expected to interact naturally with humans.
According to Meta, BlenderBot 3 features conversational skills such as personality, empathy and knowledge, incorporating long-term memory, and searching the internet to carry out meaningful conversations. These BlenderBot 3 features are also inherited from Meta’s BlenderBot 2 and built from Meta AI’s publicly available OPT-175B language model.
To avoid mimicking and generating unsafe, biased or offensive remarks, Meta stated that it has conducted large-scale studies, co-organized workshops and developed new techniques to create safeguards for BlenderBot 3. In line with these concerns, the Bot 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.
In terms of comparison with previous BlenderBots, Meta announced that BlenderBot 3 improved by 31% on conversational tasks. It’s also twice as knowledgeable, while being factually incorrect 47% less often. It was also found that only 0.16% of BlenderBot’s responses to people were flagged as rude or inappropriate.
Despite current work, Meta further mentioned that its new BlenderBot can still occasionally make incorrect, inconsistent, rude, offensive or off-topic comments.
Through the latest chatbot, Meta plans to collect and release feedback data for its use and that of the broader AI research community which can be leveraged over time. Furthermore, the company stated that it has also developed new learning algorithms to distinguish between helpful responses and harmful examples.
This announcement follows the recent launch of Meta AI’s newly built single AI model called “No Language Left Behind” (NLLB), which currently translates 200 different languages, considered as a significant breakthrough in machine translation capabilities. Meta AI researchers created NLLB as an effort to develop high-quality machine translation capabilities for most of the world’s languages.