Google CEO Sundar Pichai has announced that the language model used by Bard will be upgraded to a more powerful one, as generative AI gains momentum as a new technology frontier. The new model, called PaLM, is larger than the current LaMDA model used by Bard, with 540 billion trained parameters compared to LaMDA’s 137 billion. This upgrade will allow Bard to be more efficient in analyzing prompts.
“Pretty soon, maybe as this goes live, we will be upgrading Bard to some of our more capable PaLM models, which will bring more capabilities, be it in reasoning, coding.”said Pichai,”We clearly have more capable models,”.
It sounds like Pichai is emphasizing the importance of ensuring the quality and reliability of their technology before releasing more advanced versions. By limiting the scope of Bard with LaMDA, they were able to gather valuable feedback from a larger group of people while also keeping the required computing power low. Pichai also seems to be indicating that they are still in the early stages of development and will continue to release more capable models over time. However, he wants to prioritize getting it right rather than rushing to be the first to release the latest technology.
