Google's latest Pixel phones are the first to be powered by its Tensor SoC. There's still a lot to learn about both devices, especially regarding performance and battery life, but what we do know is pretty impressive. While Google has kept the basics of its new chip a mystery for now, it has shown off a couple of new tricks possible thanks to Tensor, including upgraded speech recognition.
According to today's Twitter thread highlighting some of the best aspects of these new Pixels, this new SoC improves voice commands, translation, captioning, and dictation. Not only could that make Google's latest phones some of the most accessible devices available today, but it could speed up tools you use every day to communicate. Simple commands, like translating a conversation while on vacation or setting reminders with Assistant, might be much improved with this chip.
Switching ⚙️ a bit. Let’s talk speech.
Speech recognition is another foundational technology where you will see a huge improvement in #Pixel6.
Google Tensor allows us to make big leaps in:
— Made By Google (@madebygoogle) August 2, 2021
In its hands-on with the device, The Verge got to see a couple of these enhancements in action. In one demo, the Pixel 6 displayed its on-device auto-captions faster and with more accuracy, even offering real-time language translations from a video. In another, voice typing recognition was near-instant, with Google hardware VP Rick Osterloh able to simultaneously dictate sentences while editing specific words with the keyboard.
Though these improvements may seem small, speeding up speech recognition — especially without an internet connection — is a big deal. It's easy to see a world where Tensor's faster responses also upgrade spaces like your car, making hands-free texting and calling better than ever through Android Auto. It's all speculation for now, but hopefully, we won't have to wait long to learn more about the Pixel 6's new chipset.