Bard is based on a language model Google built called LaMDA. After an assignment, it considers the most logical order of words for an answer. “Language models are an exciting technology, but they’re not flawless,” Google points out. For example, the information is sometimes incorrect, and the answers may reflect prejudices from the texts Bard was trained with. Google hopes that the technology will improve through many assignments from test subjects. Bard is built in such a way that users can quickly find out whether the given answer is correct in the Google search screen.
Microsoft also wants to further market its artificial intelligence. People can give the Bing platform a written assignment, after which the AI makes a drawing of it. “The brain processes visual information about 60,000 times faster than text,” Microsoft explains the importance of Bing Image Creator. The image maker works on the basis of the DALL-E technology model. That was released two years ago by researchers at OpenAI. The lab, founded by Elon Musk, among others, then published the intelligent chatbot ChatGPT. It has been in the news a lot in recent months.