Google has opened Bard, its AI chatbot tool, to the public. It directly competes against ChatGPT.
Users can sign up for a waitlist starting Tuesday to get access to Bard. This program promises to help them outline and write essays drafts, plan a friend’s baby shower and find lunch ideas based upon what’s in their fridge.
CNN was told by a company representative that it would be an entirely separate experience from Google Search. Users can also visit Search to view its results and sources. In a blog post, Google stated that it will 'thoughtfully add' the technology to search "in a deeper manner" at a later date.
Google stated that it will begin rolling out the tool in the United States, United Kingdom and other countries. It plans to expand the tool to other languages and countries in the future.
This news comes as Google and Microsoft and Facebook race to create and deploy AI-powered tools following the viral success of ChatGPT. Google also announced last week that it will be adding AI to its productivity tools such as Gmail, Sheets, and Docs. Microsoft soon announced a similar AI upgrade for its productivity tools.
Google's Bard was unveiled last month during a demo. It was later criticized for not responding to a question regarding a telescope. Google's parent company Alphabet shares fell 7.7% on that day, wiping out $100 billion of its market value.
Bard, like ChatGPT (which was made public by AI research company OpenAI in November), is built on a large-language model. These models are trained using vast amounts of online data to produce compelling responses to user requests. Google management declared a 'code-red' situation due to the immense attention ChatGPT received.
Bard's error highlighted the challenges Google and other companies have in integrating technology into core products. There are a few issues with large language models, including perpetuating biases, responding aggressively and being factually incorrect.
In a blog post on Tuesday, Google acknowledged that AI tools have 'their faults'. Google said that it uses human feedback to improve its AI tools and to add new safeguards such as limiting the number of dialogues per conversation to maintain helpful interactions.
OpenAI's GPT-4, the next generation of technology that powers ChatGPT, Microsoft's new Bing browser with similar safeguards, was released last week. GPT-4 was announced in the first day. It stunned users and company demos with its ability to create lawsuits, pass standardized tests, and build a functioning website from a hand-drawn drawing.