Here’s what we know about ‘Bard’: an artificial intelligence chatbot. It follows the success of OpenAI’s chatbot ChatGPT, which launched a test version last year. Google’s Bard will also be released as a test version, first to a small group of people and then to the public “in the coming weeks”.
What are chatbots?
Chatbots like ChatGPT and Bard are often referred to as ‘generative’ artificial intelligence because they can generate their own original responses to user questions (as opposed to predetermined ones). They can do this because they are ‘trained’ on vast amounts of online text. The bots develop a ‘large language model’ from this text that they use to answer questions.
Bard v ChatGPT:
ChatGPT’s main advantage has been in producing texts that match a general style – like an email, a recipe, a song, or a mission statement. However, it acknowledges “limitations” in providing reliable information and it is not trained to have the ability to assess truthfulness. Google’s Bard appears to focus on real-world information.
Google has previously published a series of principles for artificial intelligence, many of them related to common concerns about AI. The company has also promised not to use AI “where there is a material risk of harm” or for “surveillance violating internationally accepted norms”.
