Mycroft has open sourced Adapt, an intent parser software library, to encourage open source artificial intelligence development for IoT embedded devices.
Mycroft says it aims to assure the future of open source artificial intelligence through its release this week of Adapt, an intent parser engine for embedded devices, as an open source project.
Mycroft's main product is a device of the same name that is designed to manage IoT devices in the smart home and office. The chief selling point of the Mycroft is its ability to predict and learn what users want in an intelligent way.
That requires an AI software framework that can translate human thoughts into automated behavior. Toward that end, the company has been developing its Adapt intent parser as a library of functions that programmers can use to help devices or apps interpret natural language input.
Adapt was previously developed internally by Mycroft, but on Wednesday the company open-sourced it on GitHub. The company says that providing open source AI solutions like Adapt is crucial for ensuring that AI products remain open.
"Artificial Intelligence software like Adapt is becoming more and more important, but very little of it is open," the company said in a statement. "At Mycroft we believe that natural language processing is going to be key to the success of future products and technologies. By sharing our software we believe that we can help bring artificial intelligence to the open source community where it can be used to benefit millions of users worldwide."
That's marketing speak, of course, but there's some sound logic behind it. AI is poised to become increasingly important as IoT evolves and users demand ever-smarter smart devices. But AI software is particularly tricky and expensive to develop. That means open source coders who lack deep-pocketed supporters face an uphill battle in building open AI solutions.
As a small company, Mycroft isn't likely to be able to sustain a major open source AI initiative on its own, but its open-sourcing of Adapt is a solid first step in encouraging more programmers and companies to collaborate in order to create an open source AI code base that all vendors and developers can adapt for use in their products.