We recently reported that Google has launched a competitor to what is probably the most famous chatbot today called ChatGPT Cool AI. However, the tech giant's chatbot had certain weaknesses, specifically in the area of mathematics and logic. But that's changing now, as Google has implemented a self-developed language model into it that improves its mathematical and logical capabilities and paves the way for autonomous code generation in the future.
If you don't know, Bard is built on the LaMDA (Language Model for Dialogue Application) language model. In 2021, Google announced its long-term vision for a new Pathways model, and last year it introduced a new language model called PaLM (Pathways Language Model). And it is this model, which at the time of its introduction had 540 billion parameters, is now being combined with Bard.
PaLM's logical abilities include arithmetic, semantic parsing, summarization, logical inference, logical reasoning, pattern recognition, translation, understanding physics, and even explaining jokes. Google says Bard can now better answer multi-step word and math problems and is soon to be enhanced to be able to generate code autonomously.
You could be interested in
Thanks to these abilities, in the future Bard could become an assistant to (not only) every student in solving complex mathematical or logical tasks. Anyway, Bard is still in early access in the US and UK at the moment. However, Google has previously said that it intends to expand its availability to other countries, so we can hope that we will be able to test its mathematical, logical and other abilities here as well.
