The Google chatbot Bard was the first to be tested in the United States and the United Kingdom in March.

(Photo/Reuters)

OpenAI, an American start-up company, launched the ChatGPT conversational chatbot built with generative AI artificial intelligence technology at the end of last year. Since the opening of the beta version of the official website, it has become popular all over the world, and it has set off competition among major technology giants such as Microsoft and Google. AI war.

Google released the self-developed conversational robot "Bard" in February, and launched the initial public beta in the United States and the United Kingdom in the middle and late March (currently only supports the English version).

According to foreign media New York Times, Google CEO Sundar Pichai said in an interview recently that an "upgraded version" of Bard will be launched in the near future, which will have stronger reasoning and mathematical logic capabilities.

Please read on...

(Photo/Reuters)

The biggest difference from Bard, which has previously opened a public beta experience, is that Bard in the initial stage of launch is based on the lightweight LaMDA language model training center.

The Bard, which is about to usher in an "upgraded version", will be trained and operated based on a larger-scale PaLM language model.

The former LaMDA model uses a language database of 137 billion parameters, while the latter PaLM model has a language database of up to 540 billion parameters. The powerful computing mechanism makes the content naturally produced by AI evolve with it, bringing more understanding of human speech. Language comprehension, multiple reasoning, and problem-solving skills.

The reason for taking a step-by-step approach, Sundar Pichai pointed out that the main consideration is that Google hopes to focus Bard, which has the most advanced AI dialogue technology, on providing more natural interaction, and more accurate and secure information content.

It is worth noting that foreign media CNBC reported last week that Google is currently undergoing a period of major personnel changes and organizational reorganization.

It is reported that the content of the internal memo sent to employees mentions the integration of the Google Assistant voice assistant department and the Bard development team, and emphasizes that "Bard" will be the key axis of the department in the future to bring greater impact to users force.

Jianchang Mao, the vice president of engineering who was in charge of Google's voice assistant department, will leave due to personal reasons. The position will be taken over by Peeyush Ranjan, the vice president who was originally in charge of Google Pay payment business.

In addition, Amar Subramanya, the director of engineering research and development who was originally in charge of Google's voice assistant department, will lead the development of "Bard" in the future.

To put it simply, the voice-activated Ms. (or Mr.) of Google, which is familiar to Android users, was previously in charge of the Google Assistant voice assistant department, which implemented the voice-activated application technology to various software and hardware devices owned by Google (such as: mobile phones, smart watches, car driving, etc.).

After this wave of personnel changes and organizational structure adjustment strategies, what kind of division of labor will be adopted between the Google voice-activated assistant service and the upgraded version of the AI ​​​​chat robot "Bard" in the future? Imported into the scope of Google Assistant-related applications, it will still develop in the direction of integration, for example: training Miss Google Voice to transform into "Bard" has the ability to generate powerful natural language content, or let "Bard" have the ability to execute Miss Google Voice Voice control capabilities, these are the focus of high attention from the outside world.