Let’s welcome our friends at Deepgram to This Week In Voice VIP as our Presenting Sponsor. Thanks to them for their partnership and support. Phenomenal company.
If you don’t know who Deepgram is or what they do, take a moment to learn.
AC/DC’s Back in Black is one of the greatest rock songs of all time, spawning covers with over 15 million views and genius mashups like this.
The Guardian interestingly described the song’s “judicious use of space” as singularly important in the song becoming a defining anthem.
Fascinatingly, what is arguably the band’s signature song didn’t get written until the band’s seventh album was released, and only in response to one of the band’s original members passing away. The song was, in many ways, a rebirth.
While in Austin, Texas, a week or so ago for The Voice of Gaming event, I had a chance to meet with Eiman Ebrahimi, a decade long research scientist at NVIDIA who recently stepped out on his own about a year ago to join UCSD Prof Hadi Esmaeilzadeh who had taken leave from UCSD to start Protopia AI.
Protopia, which features the tagline “Empowering Responsible AI,” takes data streams of either audio, video, or both, and transforms them to remove unnecessary features and information and thereby make them simultaneously more manageable and private.
I sat with Eiman for about an hour, discussing with this University of Texas PhD (who got his undergraduate degree in Tehran) ins and outs of data manipulations, which he certainly understands much more than I do.
“Why do we continue to think about AI ‘listening’ and ‘viewing’ sound and video in the same way we do? A computer’s senses aren’t the same as human senses and we need algorithms to adjust for that,” Ebrahimi said.
But Protopia feels like the right play: letting privacy and security enhancements of personal data naturally occur as byproducts of optimization processes, rather than some sort of primary, intentional act.
The company focuses on video in its marketing, but audio and voice is a major area of emphasis for the company. The need to increase speed of voice assistants by locally reducing voice queries and commands down to optimized bare essentials exists for every platform right now.
Well worth keeping an eye on these folks.
Also in the aftermath of Austin’s gaming event, I spoke with Wanderword’s Henrik Lindfors - lead designer for the leading voice gaming company on the planet.
We spoke about gaming - a hobby of both of ours - and what has happened to pave the way for voice gaming to now be a viable business, as opposed to being much more difficult just a few short years ago. The video is available here.
In addition to partnering with our friends at Deepgram, who are now the Presenting Sponsor of This Week In Voice VIP, I have made the decision to permanently make all newsletter content free.
The paid tier will remain, though, and those paying into it will not only show appreciation for the newsletter, but more importantly, that revenue will directly subsidize complimentary Project Voice Series conference registrations for those in financial need.
National Geographic.
Verizon.
The Washington Post.
ReadSpeaker. Rabbit & Pork. Canary Speech. Atexto. HEAD acoustics.
These and many other companies will be in person at Project Voice 2022, the #1 event for voice/AI in America.
Condensed single-track program, creating more space for social activities and networking.
Capped at 750 executives, in the Chattanooga Convention Center, which is home to a unique air circulation system that makes indoor air like the outdoors. We’ve done multiple events there safely the last two years and this will be yet another one.
Hope to see you there!