Google announced access to Bard, an artificial intelligence chatbot powered by a large language model, on Tuesday (LLM). Bard is Google’s response to OpenAI’s ChatGPT, whose popularity has skyrocketed since it went public late last year.

The company wrote in a blog post, “We’ve learned a lot so far by testing Bard, and the next critical step in improving it is to get feedback from more people.”

Google has proven to be more cautious than Microsoft’s OpenAI in the tech giants’ competition to dominate the AI industry. While ChatGPT has become the current buzzword for artificial intelligence, Google has struggled to generate the same enthusiasm.

With the decision to make Bard public, Google is no longer on the sidelines, but it is also not abandoning caution.

The company wrote, “While LLMs are an exciting technology, they are not perfect.” “Due to the fact that they learn from a wide variety of data that reflects real-world biases and stereotypes, these tend to manifest in their outputs. In addition, they can confidently present inaccurate, misleading, or false information.”

This is precisely what occurred when Google introduced the tool at a press conference. It provided the incorrect response to a question about a major scientific achievement, resulting in an 8 percent drop in the stock price of Alphabet, Google’s parent company.

Google added that the testing process is designed to avoid these pitfalls by identifying where guardrails must be constructed to ensure that interactions are “helpful and on topic.”