Skip to Content

An Open-Source (and Cute) Alternative to Amazon Echo

Mycroft’s voice-enabled assistant is embodied in the startup’s friendly looking Mark 1 gadget, but you can build it into all kinds of devices.
MYCROFT

Mark 1 is no Amazon Echo: it looks like an ’80s clock radio mashed up with WALL-E, and speaks with a robotic, bass-heavy British accent. But the startup behind it, Mycroft, hopes it has similar appeal to hackers, students, and companies who want a voice-enabled assistant that they can run on all kinds of devices and alter at will.

When it comes to voice-enabled digital assistants, there are plenty of them available these days—in addition to the Echo, which runs Amazon’s Alexa assistant, there’s Apple’s Siri, Microsoft’s Cortana, and Google’s assistant. None of these is open-source, though, so even if developers can use it on various devices (like Amazon’s Alexa), they can’t go under the hood and change its code—ostensibly, to help improve it.

Mycroft—whose voice assistant, which runs on Mark 1, is also called Mycroft—isn’t trying to rival any of these big companies’ digital helpers, says CEO Joshua Montgomery. Rather, he says, the idea is to democratize the voice assistant—making it available and adaptable for everyone from kids working on school projects to companies that want to use open-source voice-enabled technology for a call center. While Mycroft will be free to individual users like consumers and developers, who can download it to run on things like computers and other devices, the company plans to charge enterprise users.

Mark 1’s LED face has a friendly look.

Mark 1, which is for sale online for $180, is a physical manifestation of what that can look like.

“We decided to give it a face and to make it cute and personable so people would use it,” Montgomery says.

So far, there are about 90 skills users can activate for Mycroft; the company came up with about 16 of them and developers contributed the rest. They include things like setting reminders, playing music, making Facebook posts, and controlling lights and other connected devices at home. The device comes with some skills already installed, Montgomery says, and users can then add more.

Montgomery showed off some of Mark 1’s capabilities for me, alerting it by saying “Hey, Mycroft”—to which a wavy LED line showed up on its face, as a sort of mouth—and then asking it a range of questions like “Who won the 1959 World Series?” and “How’s the weather in San Francisco?”

Most of the time, it answered correctly, using sources like Wikipedia and Wolfram Alpha to gather information. It was a little sluggish to respond, though, and it didn’t always do as it was told. The company tries to make it clear that it’s still a work in progress, and a more consumer-friendly version of Mycroft with greater capabilities is slated for next February. It can be set to use one of several different speech-to-text systems, including that of Google or IBM—and Mycroft is building its own open-source version, called OpenSTT.

The company hopes Mycroft will fit in outside of the home, too. Car maker Jaguar Land Rover made a strategic investment in the company and plans to get its technology into vehicles within several years, Montgomery says.

Carolina Milanesi, principal analyst covering consumer technology for tech market researcher Creative Strategies, thinks bringing Mycroft to the car or using it to imbue various home gadgets with voice interactions could be a good tactic, rather than trying to compete with much bigger players in the digital-assistant realm.

There are a number of important issues to consider, though, such as how secure the data is that Mycroft is able to access.

“For a lot of these interfaces, you’re starting to control very important parts of your life, from the home to the car, and these things need to be secure,” she notes.

Montgomery says that user data sent to Mycroft will only be used by the company for improving its technology if users give permission to do so. This could make it tricky for Mycroft’s software to get better over time, though, as the more data it can draw on, the smarter it can get.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.