Google’s Bard is singing a new tune this week, upgrades to its capabilities already rolling out

For anyone who has access to Google’s Bard AI, it’s evident that it is losing the chatbot race against ChatGPT and Bing AI (never thought I would say that!). It gets a bunch of very basic things wrong and is also the only AI that’s not sure if it’s sentient. Google has acknowledged that Bard is in early beta development and will make mistakes. However, critics have pointed out how it pales in comparison to its competitors. While great for some things, especially since it’s trained on Google Search, it’s just not up to many tasks yet.

Well, that is until now. After being accused of using data from ChatGPT directly, Google and Alphabet CEO Sundar Pichai has stated in an interview on The New York Times’ Hard Fork podcast (kudos: The Verge) that “Pretty soon, perhaps as the [podcast] 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, or answering math questions.” He also stated that users would see progress over the course of the next week. It’s very likely that it’s already receiving these upgrades, and some individuals in the Bard Discord have already pointed this out.

Apparently, Bard itself has been running on a “lightweight and efficient version of LaMDA,” which is Google’s controversial language model. The company has basically limited Bard’s capabilities to ensure consumer safety, as confirmed by Sundar during the aforementioned podcast. However, with the upgrade, Bard will receive a more powerful version of LaMDA that will inject it with more intelligence and expand its capabilities.

Lastly, he explained that it was important not to release a more capable model until Google could ensure that it was safe for consumers. Google has repeatedly chimed in that it’s committed to responsible AI development and ensuring that its models are secure. This week, the industry has proposed a freeze on new AI models being released since they’re so powerful, and research experts have agreed to assess concerns before going crazy with this new AI revolution.