Are cloud-based voice assistants doomed? That seems like an overly dramatic question to ask if you look at the current state of the millions of Google Assistant, Amazon Alexa, and Apple Siri users, but we’re not so sure about the future. Google and Amazon have backed their voice assistants recently, with Amazon firing much of the Alexa team due to losing $10 billion a year. Google isn’t quite at the “fire for everyone” stage, but it’s reportedly less interested in supporting the Assistant on third-party devices, which would be a terrible move given Google’s very small hardware division. Everyone built these systems assuming that a stream of income would come later, but that income never came, and it looks like the bubble is bursting.
One project that relies heavily on Big Tech voice assistants is sitting around and waiting for doomsday. The team at Home Helper are declaring 2023 “the year of the Home Helper Voice.” This is basically the smart home flagship saying, “If these cloud voice assistants don’t provide a multi-billion dollar revenue stream for Big Tech, that’s fine, we’ll do it ourselves!” There are already several new open source voice assistant projects, but the Home Assistants team has proven that they can manage a large project. It has a large, thriving community and enough revenue to support full-time employees, making it the new frontier for viable local voice service.
Plus Home Assistant isn’t starting from scratch – it went and found the “most promising” open source voice assistant out there, “Rhasspy,” and hired its lead developer, Mike Hansen, to work full-time on voice i. Home Assistant. Hasen will now work at Nabu Casa, a Home Assistant commercialization company. According to Home Assistant founder Paulus Schoutsen, “Rhasspy stands out from other open source voice projects because Mike doesn’t just focus on English. Instead, his goal is to make it work for everyone. This is going great as Rhasspy already supports 16. different languages today.” It is planned to support all 62 languages currently supported by the Home Assistant, but with voice, all without the need for an Internet connection.
Schoutsen says that Home Assistant will keep the scope of the project “manageable” for now and will “limit the number of possible actions and focus on the basics of interacting with your smart home. There are no web searches, making calls or voice games. ” On of course, there’s probably some way the Home Assistant community can tack on extra features to it, and then it’ll quickly emulate the Hal 9000 and do a million more things.
The problem with all these Big Tech voice services is that they don’t have a way to generate ongoing revenue. They don’t really have a way to show ads, and no one wants another subscription service. However, they generate an ongoing cost due to the server time required to process all that voice communication. Google and Amazon compounded the problem by selling their voice hardware at cost in an attempt to win the voice assistant rush, hoping for an additional revenue stream later. Apple launched a high-end Siri speaker, the HomePod, in 2018 at a staggering $350 price tag, but in retrospect, it looks a lot more sustainable than whatever Amazon and Google were doing.
Google seems to have tried to solve this problem with its second-generation voice speakers, which moved “some” voice processing to local chips. Moving some voice processing off the cloud will reduce server time, but it’s unclear if that’s enough to meet Google’s incredibly high standards for ongoing product support.
By moving a voice assistant into a self-hosted environment, you can afford those costs at a reasonable rate (some sort of hardware to start with and then just the electricity bill) without worrying about the weird, shifting priorities of Big Tech. ecosystem. There are likely to be many other benefits as well. A local voice assistant would be great for privacy, and it will feel less like you’re always running a wiretap in your home 24/7. Home Assistant has also proven to keep all these local results in a much faster experience for your smart home, and Google has said the same about its limited local voice processing.
The Home Assistant already has a text-only “Chat” command system, so it’s a matter of taking that out and wiring it up to voice input and output. What is not clear about this is the hardware solution beyond the “pile of wires and circuit boards” that usually dominates open source smart home projects. Part of what makes Google Assistant and Alexa so popular is a selection of good-looking speaker/microphone mates that you can spread around the house without looking like a mad scientist. Home Assistant commercialization company Nabu Casa has been working to close this hardware gap lately, making plug-and-play server boxes and dongles to bring things like Matter support into the Home Assistant ecosystem. Maybe he could send a speaker.