
What is the fate of Google Assistant? Evidence is mounting that the sector is crumbling. The latest news comes from CNBC’s Jennifer Elias, who says the Google Assistant division has been “reshuffled” to “hugely prioritize” Bard over the Google Assistant. It looks like the team has been reassigned.
We’ll get into the details of the report shortly, but first, here’s a quick rundown of what the Assistant has been through under Google over the past two years.
- Google Assistant saw eight major speaker/smart display hardware releases in the five-year period from 2016 to 2021, but hardware releases seem to have stalled. The last hardware release was March 2021. That was two full years ago.
- We saw Google in 2022 delete Assisted support with two in-house product lines: Nest Wi-Fi and Fitbit wearables.
- In 2022, there’s also a report from The Information that says Google will “reduce investment in developing voice search for cars and non-Google devices with the Google Assistant.”
- Google Assistant driving mode will be retired in 2022.
- Google Assistant’s “Duplex on the web” feature will also be retired in 2022.
- One of Google Assistant’s core and unique features, Reminders, will soon be retired in favor of Google Task Reminders.
- Google Assistant has never made money. The hardware is sold at a fixed price, there are no ads, and no one pays a monthly fee to use Assistant. And while processing all those voice commands comes at a significant server cost, some new devices are moving to on-device processing in a stealthy cost-cutting move. , Amazon Alexa, is in the same boat, losing $10 billion a year.
Each of these developments may be dismissed individually, but together they begin to paint the familiar picture of Google’s looming shutdown.
According to the latest news from CNBC, it looks like the assistant team won’t be working on assistants anymore. A memo to employees titled “Assistant and Bard Team Changes” lists a number of executive changes. Amar Subramanya, Google Assistant’s VP of Engineering, will lead engineering for his Bard. Another Google Assistant VP of Engineering, Jianchang Mao, is leaving Google “for personal reasons.” Mao will be replaced by Peeyush Ranjan, who is currently vice president of Google’s commerce division and oversees payments. (Payments at Google have been an incredible disaster over the past few years. Seeing someone run away and run another division raises eyebrows.)
The memo, from Sissie Hsiao, vice president and head of the Google Assistant business unit, instructs the Assistant team to: [their] We want to support future opportunities and make sure we continue to perform. The assistant he team now appears to have a role in helping Google’s “Code Red” fight against ChatGPT.
Integrating Google Assistant with Bard makes some sense, if the two units aren’t separate products. Everyone expected Bard to somehow integrate with Google Search, like ChatGPT and Bing, but the product the company released was a standalone “experimental” that wasn’t connected to search at all. It’s a chatbot. Just like ChatGPT, it can generate paragraphs of suspiciously accurate text based on what you see on the Internet.
Google Assistant is a voice product that is primarily concerned with the accuracy of speech recognition, recognizing and performing voice tasks such as “turn off the lights”, “set a reminder”, “ambient computing”, or performing various tasks. Available anywhere on your device. Star Trek computer. Although there is some overlap as both services can return answers, Google Assistant’s current Google Search powered answer system is a perfect fit for Google Assistant. Google Bard can generate paragraphs of text, but when those answers are read aloud, the Assistant’s short answers are better than over and over again with a clumsy, monotonous text-to-speech system.
Bard’s and Assistant’s interfaces are similar in that they both look like chat apps, so they have similar monetization issues, but they’re actually using both products for different purposes. Assuming that the idea of the Google Assistant (a voice assistant that helps users interact) isn’t completely dead at Google, we can imagine a future where Bard’s language model can understand what users want and help them do it. Today’s assistants don’t have language model issues, just voice recognition issues, and Bird won’t help with that.
We’ve seen Google do this all the time. That is, shutting down a good project in favor of an alternative project that isn’t ready or is riddled with feature regressions. Bard’s development (and Google’s AI strategy in general) seems slow compared to the rollout of OpenAI and Microsoft’s ChatGPT, so perhaps more people will help it. Not that you need it, but the service is purely a money pit.