Last week, the Linux Foundation created a new open source industry association called the Open Voice Network (OVN). The new group is an independently governed directed fund of the Linux Foundation, with the goals of improving trust, choice, inclusivity, and openness in voice recognition technology.
While similar to Amazon’s Voice Interoperability Initiative in some ways, the OVN claims a primary focus on ethics. Although Amazon’s effort does touch on user choice and freedom, its primary goal is considerably narrower than OVN’s—to provide “multiple, simultaneous voice services on the same product, each with its own wake word.” Amazon did not say much about ethical restrictions or guidelines for those individual services, either in its current mission statement or in the 2019 press release announcing it.
The Open Voice Network is a neutral nonprofit industry association with some impressive names in its founding-member registry, including Target and Microsoft. OVN’s goal is not to develop the technology itself but to deliver open, trustworthy, and inclusive standards and usage guidelines.
OVN’s initial press release goes into further detail, stating that the association will begin by focusing in these areas:
- Standards development: research and recommendations toward the global standards that will enable user choice, inclusivity, and trust.
- Industry value and awareness: identification and sharing of conversational AI best practices that are both horizontal and specific to vertical industries, serving as the source of insight and value for voice assistance.
- Advocacy: working with and through existing industry associations on relevant regulatory and legislative issues, including those of data privacy.
Membership in the Open Voice Network entails supporting its research, awareness, and advocacy with direct resource contribution as well as active participation in its conferences and workshops.
Mike Dolan, senior VP and project manager at the Linux Foundation, says that “voice [command] is expected to be a primary interface to the digital world… it is already increasingly being used beyond smart speakers to include applications in automobiles, smartphones, and home electronics devices of all types.” He goes on to say that “the potential impact of [voice] is staggering, and we’re excited to bring it under the open governance model of the Linux Foundation.”
Target VP of Architecture Joel Crabb describes Target as “continuously exploring and embracing new technologies,” including voice recognition, and states that “the Linux Foundation, with its role in advancing open source for all, is the perfect home for this initiative.”
Microsoft Azure AI General Manager Ali Dalloul underlines the importance of the technology itself, saying, “to speak is human”—and following that up with a bold statement that “voice is rapidly becoming the primary interaction modality between users and their devices and services at home and work.” (We aren’t certain we agree that voice command as the primary method of communication with devices and systems will arrive soon—or perhaps ever—but it’s certainly increasingly important.)
Founding OVN member and Schwarz Gruppe Chief Digital Officer Rolf Schumann points out the stark need for ethical design and standards in voice recognition: “voice includes more information than a fingerprint and can entail data about the emotional state or mental health of a user. Therefore, it is of utmost importance to put data protection standards in place to protect the user’s privacy. This is the only way we will contribute to the future of voice.”
Sponsoring memberships of the Open Voice Network are available to enterprises in Platinum, Gold, and “Advocate” levels at $100K, $50K, and $7.5K annually, respectively, with Platinum sponsors serving on the OVN steering committee and the other two levels “represented” on the committee. Individuals and smaller businesses wishing to get involved may do so as Friends of the Open Voice Network—a commitment that requires volunteer participation and signing a charter but no cash contributions.
https://arstechnica.com/?p=1776791