Home » News » Google Assistant Makes Better Of Siri and Alexa In Accuracy Test
Google Assistant Makes Better Of Siri and Alexa In Accuracy Test
December 26, 2022 By Jozeph P
(Image Credit Google)
Google Assistant has long been regarded as the best voice assistant on a smartphone, particularly in terms of accuracy and overall functionality. And it's not surprising given that Assistant has full access to Google's vast knowledge and models used to train its AI.Bespoken, an AI training company, recently put Google, Apple, and Amazon's voice assistants through their paces using an Amazon Echo Show 5 (Alexa), an Apple iPad mini (Siri), and a Google Nest Home Hub (Assistant).According to Voicebot.ai, Bespoken asked the questions using a test robot and labeled them as complex or simple. Some questions, such as "name the first man on Mars," were designed to fool the digital assistants because there was no correct answer.Surprisingly, the percentage of questions correctly answered by Google Assistant, Alexa, and Siri was roughly the same regardless of whether they were classified as complex or simple. This suggests that incorrect answers were not caused by the digital assistants' failure to understand the questions, but rather by their lack of knowledge of the correct answers. Google Assistant answers simple questions correctly 76.57% of the time, Alexa 56.29% of the time, and Siri 47.29% of the time. Google Assistant correctly answered 70,18% of complex questions, compared to Alexa's 55.05% and Siri's 41.32%.Bespoken's Chief Evangelist Emerson Sklar commented on the results of the test and stated, "We have two major takeaways from this initial research. First, while Google Assistant outperformed Alexa and Siri in every category, all three have significant room for improvement. These results underscore the need for developers to thoroughly test, train, and optimize every app they build for these voice platforms."Sklar added that "second, this process was completely automated, and we plan to continue to run these tests as well as introduce new benchmarks. This sort of automation enables not just point-in-time measurement but also continuous optimization and improvement. We know Google, Amazon, and Apple all embrace this, and we encourage others to as well."
