Bard, a new experimental AI chatbot product from Google that directly competes with ChatGPT, is now available to the public. Bard, which enables users to arrange a friend’s baby shower, prepare essay outlines and drafts, and come up with lunch ideas based on what’s in the fridge, has a waitlist that users can join to gain access. According to a company official, it will be a different, complimentary experience to Google Search, and users may still access Search to verify its conclusions or sources.
In a blog post, Google stated that it intends to “thoughtfully” add substantial language models to search “in a deeper way” in the future. In addition to the United States and the United Kingdom, Google said it would eventually roll out the capability to other nations and languages. The information is released as Google, Microsoft, Facebook, and other tech firms scramble to create and implement AI-powered solutions in the wake of ChatGPT’s recent viral popularity. The addition of AI to Google’s productivity tools, such as Gmail, Sheets, and Documents, was revealed last week.
Microsoft soon after unveiled a comparable AI enhancement for its productivity products. Then, last month, Google revealed Bard in a demo that received criticism for incorrectly answering a question regarding a telescope. That day, Alphabet, the parent company of Google, saw a 7.7% decline in share price, erasing $100 billion from its market worth.
Yet, Bard’s error made clear the difficulty Google and other businesses have incorporating the technology into their primary services. Big language models can have problems, like reproducing prejudices, being factually incorrect, and reacting violently.
In the blog post, Google admitted that AI tools are “not without their shortcomings” The business claimed that it still uses user feedback to enhance its systems and introduce fresh “guardrails, including capping the number of exchanges in a chat, to try to keep interactions beneficial and on topic.”
Bard is based on a sizable language model, just like ChatGPT, which OpenAI, an AI research organization, made available to the public in late November. These algorithms are trained on enormous web data sets to produce persuading responses to user inquiries. According to reports, Google’s management declared a “code red” situation for its search company due to the enormous attention paid to ChatGPT.
GPT-4, the most recent iteration of the technology that powers ChatGPT and Microsoft’s brand-new Bing browser, was made available last week by OpenAI with comparable security measures. The ability of GPT-4 to file lawsuits, pass standardized tests, and create a functioning website from a hand-drawn design astounded many users in early trials and a company demo on the first day after it was released.