Mashable: Google's Language Model for Dialogue Applications (LaMDA) is said to be what gives the bot its ability to talk in a natural way.

The race to the top in the field of AI is getting intense. Skynet from The Terminator is not yet a reality, but chatbots are. Reports have surfaced that a team of Google researchers is hard at work on what is widely seen as the company's reaction to OpenAI's intelligent chatbot, ChatGPT.

Apprentice Bard, a chatbot being developed by Google based on the company's own big neural language model, LaMDA (Language Model for Dialogue Application), was featured in a CNBC article. Apprentice Bard appears to be the basis for Google's efforts to redesign its current homepage, which features a prominent search bar at the top.

This comes just a few weeks after the Internet giant's management reportedly sent out a "code red" about ChatGPT, which has been in the news lately because it can write the text that looks like it was written by a person.

According to CNBC, users of Apprentice Bard can ask questions in a dialogue box and receive answers in a similar fashion to ChatGPT. The user community has the option of commenting on the bot's response.

Google's head of AI at the time, Jeff Dean, warned staff that the business is "more conservatively run than a small startup" because the "reputational risk" of supplying incorrect information is much larger.

However, both he and Pichai hinted that Google will make such products available to the general public in 2023. Google has been a leader in artificial intelligence (AI) for a long time, but web search is still its main business.

Perhaps not unexpectedly, Google teams have also been pitting a pre-release version of LaMDA chat against ChatGPT. For the purpose of doing side-by-side comparisons, it chose some sample questions and responses and placed them in their own papers.

One comment in an internal document comparing LaMDA to ChatGPT notes, "Amazingly, ChatGPT is hired at L3 when interviewed for a coding position." It didn't say if LaMDA would have done as well or not.