Google is juicing up its AI chatbot Bard with some extra horsepower – the company’s most capable AI model yet, Gemini. This bump in performance comes on the heels of reports that claimed Gemini’s launch was being delayed over multilingual issues. But Google seemingly felt pressure to push out its ChatGPT rival ASAP.
Gemini comes in three sizes, but we’ll focus on the two beefiest versions that directly impact Bard – Pro and Ultra. According to Google, Pro can handle a wide variety of tasks and will power other Google products too. Importantly, the company fine-tuned Gemini Pro specifically for Bard, improving its abilities around understanding questions, summarizing information, planning, and reasoning.
To showcase the upgrade, Google had scientist YouTuber Mark Rober ask Bard to design “the most accurate paper airplane.” Bard then provided various optimized paper plane schematics, demonstrating how it can now play a bigger creative role.
Bard with Gemini Pro is live now in over 170 countries, available in English only for text queries. Google intends to later expand support to more European countries and languages.
Early next year, Bard will get another upgrade called Bard Advanced, this time using the top-shelf Gemini Ultra model for “highly complex tasks.” Ultra features strengthened reasoning skills to carefully tackle tougher questions.
Google is safety testing this new version before launch. When ready, Bard Advanced will handle text plus accept multimodal inputs like video and code. A select group of testers will get early access, though Google hasn’t shared specifics yet on how to participate.
So while Bard may have gotten off to a rocky start, Google is now funneling some serious AI muscle into the chatbot with its Gemini tech. If these upgrades deliver, Bard could soon shape up into a formidable ChatGPT competitor.