Google's CEO, Sundar Pichai, has announced upcoming upgrades for the experimental AI chatbot Bard, after its failed demo in February, reported first by Interesting Engineering.

Pathways Language Model

In an interview for The New York Times' Hard Fork podcast on Friday, Pichai "assured" listeners that Bard would soon be upgraded to a more capable Pathways Language Model (PaLM).

He discussed the limitations of the current version of the Bard AI chatbot, which is operating on a lightweight and efficient version of LaMDA, and how it will be upgraded to a more capable Pathways Language Model (PaLM).

He further explained that PaLM is a more advanced language model that is better suited to handle tasks like logical reasoning and coding issues, and the upgrade will bring more capabilities, including better answers to math questions and improved reasoning and coding abilities.

 

On March 21, Bard was made available to the general public in the US and the UK, but it has not drawn the same interest as OpenAI's ChatGPT and Microsoft's Bing AI chatbot, which boasts millions of users.

Pichai acknowledged that Bard's limitations were a result of Google's perceived conservatism and that he had discussed the project with Larry Page and Sergey Brin, two of Google's co-founders. He emphasized that it was crucial to ensure that a more capable model could be handled before releasing it. 

Pichai insisted that he did not initiate any "code red" to halt development, but it is unquestionable that others in the organization received emails indicating there is a code red. 

He also discussed the need for regulation in the current industries and expressed concern about the dangers brought on by the quick development of AI. According to him, whether or not AGI has been attained hardly matters because AI systems will be incredibly powerful.
 
AGI or Artificial General Intelligence refers to a hypothetical machine that possesses intelligence comparable to that of a human being.

AGI would be capable of understanding or learning any intellectual task that a human being is capable of performing. 

Improving Bard

Bard will soon receive enhancements that will increase its functionality and competitiveness compared to its current state.

Pichai claims that Bard's reasoning, coding, and math problem-solving skills will all be enhanced by PaLM's new features. The upgrades should be finished soon although there is no official date announced yet. 

Google will have to bounce back from a mistake that the Bard AI chatbot committed last February. During a live-streamed event, Bard was asked about discoveries from the James Webb Space Telescope that a 9-year-old kid could learn about.

The chatbot provided an incorrect answer after claiming that Webb took the first-ever images of exoplanets even though it was the Very Large Telescope of the European Southern Observatory that accomplished such a feat in 2004. 

The error by Google cost its parent firm, Alphabet Inc., $100 billion in market value after that incident. 

 

Source - Tech Times