Quite some time ago, the Mycroft personal assistant was released. Unfortunately, that initial released failed to gain much traction. To many in the open source world that was a surprise, considering it was possible to install Mycroft on a desktop without having to buy the hardware, although it was quite the challenge. In fact, you could get the Mycroft experience on a number of platforms.
That first release targeted more toward developers and enthusiasts showed much promise. It was rough around the edges, but the spirit of open source was thick in the project and Mycroft offered something its competitors couldn’t: Privacy.
The court of public opinion is still split on the likes of the Amazon Echo and Google Home. Why? Because they listen and record. How much they record and for how long those recordings are retained is a matter of speculation. Even so, anyone with an eye on their privacy would find both products suspect.
That’s where Mycroft Mark II comes in to play (Figure A).
This evolution of the original, which targets consumers, looks seriously promising. Currently still in its Kickstarter phase, though it did reach its funding goal in 6.5 hours, the Mycroft Mark II is a consumer ready smart speaker that offers a built-in screen, optional camera, and state of the art microphone array with noise cancellation and beam-forming. Mycroft Mark II is a wireless smart speaker that can:
- Play music
- Set timers
- Access calendar events
- Search for general knowledge
In other words, Mycroft can do everything the public has come to expect from a state-of-the-art voice assistant. But to help differentiate itself from Echo and Home, Mycroft offers:
- Privacy, by deleting queries as they come in.
- The ability to customize the wake word, voice, and even the user experience.
- A neutral platform. Mycroft represents the user, not a company.
- Mycroft uses data (only from users who opt in to publish an open data set) which can be used to improve wake word spotting, speech-to-text transcription, natural language understanding, and speech synthesis.
- A completely open platform (software, hardware, and data).
SEE: Information security policy (Tech Pro Research)
For me, the Mycroft Mark II represents something neither Echo or Home can possibly offer. Openness. At home, we have two Echos, and everyday I wish those devices could better integrate with other technologies. In some cases it can, but in most it’s a task that ends in disappointment. Don’t get me wrong, the Amazon Echo is an incredible product–but when compared to the prospect of Mycroft, it falls slightly short.
With the Mycroft Mark II on the horizon, developers of all kinds can work their magic to create new abilities and integrate the product with other software and services. The possibilities are seemingly endless.
Imagine this: You’re in a room where a Mycroft Mark II device awaits your command. You’ve managed to integrate Mycroft to your desktop, such that it can work as a speech-to-text tool and save files in .odt format. You get an idea, open up a new file, and start dictating to Mycroft. Your dictation is saved on your desktop, in a format that can then be opened in LibreOffice.
With Mycroft, that could be a possibility. In fact, just about anything is possible with this open source project. Even better, with Mycroft, your dictated ideas won’t be saved on a third-party server. It’s all private. Your conversation with Mycroft will not be sold to companies as data to be used for targeted advertisement. When you ask Mycroft about something, said something won’t pop up on your Facebook or Twitter feed seconds later.
However, to be perfectly clear, Mycroft does send your data to a cloud server to process requests. The difference between Mycroft and Echo or Home is that the requests are immediately deleted. The Mycroft developers say they never save data or use data for advertising purposes unless you opt in (at which point, you may as well be working with either Amazon’s or Google’s product).
The big win for consumers is Mycroft’s the default action to not save data. Instead of an opt out, Mycroft is an opt in. If you want your data saved, you can configure Mycroft to do just that. Otherwise your data will not be saved or used.
As for companies? Your business could make use of Mycroft (in many clever ways … thanks to it being fully open) without the concerns for privacy found in Echo or Home. Or, better yet, your business could integrate Mycroft into a project to further expand its abilities. In fact, you could create a startup, based on bridging Mycroft with another product, service, or server. That’s that beauty of this project being completely open.
It’ll be a slow start
I suspect the Mycroft Mark II will have a slow start out of the gate. The price is on point. For a $129.00 pledge, you can reserve a Mycroft Mark II (estimated shipping is December, 2018), which is on par with the competition. It will take some doing to gain traction in a market dominated by Amazon and Google, but once the public gets wind of a competitor that doesn’t record and save their commands (and who knows what else), Mycroft should take off. As long as the initial releases are solid and don’t require much in the way of skills to set up, this could be the next big thing in personal assistants. I only hope the company behind Mycroft is prepped for a deluge of orders, once word gets out.