Google made its AI chatbot available to the general public

 Google made its AI chatbot available to the general public.

Google has launched its own conversational robot named Bard, aimed at improving the quality of responses through greater interaction with users. The launch comes five months after the successful release of the ChatGPT interface by OpenAI in collaboration with Microsoft. The creation of Bard was announced in February as a reaction to ChatGPT's appearance in November. Although initially limited to "trusted testers," Google has now made it available to the public with certain restrictions. Access to Bard will be limited to users on a waiting list in the United States and the United Kingdom for the next few months. Bard is an independent interface on a website with a text window for users to ask questions of all kinds. Google aims to receive feedback from more people to improve Bard. The vice presidents of Google, Sissie Hsiao and Eli Collins, shared this information on their official site.

According to Google executives Sissie Hsiao and Eli Collins, the more people use Bard, the better Long Language Models (LLMs) will be at predicting useful responses. Bard is built on LaMDA, a language model created by Google to generate chatbots, and Google has implemented safeguards to prevent it from delivering inaccurate or inappropriate responses, including limiting the interaction duration. Experts predict that search engines may disappear as AI-based language models become more prevalent. While these tools have benefits such as providing instant responses to user queries, they also have potential drawbacks. With generative AI, these systems can provide coherent text and answer a wide variety of questions as long as the user specifies what they want.

ChatGPT and Bard are capable of providing personalized responses, thanks to machine learning techniques trained on text available in publications such as Wikipedia and PubMed. However, experts point out that these language models often produce "nearly repeated statements" without true understanding, and what appears to be artificial intelligence is simply a reflection of what is written and trained by humans. Even if a language model system produces few errors, humans may not be able to discern them beforehand.