View profile

Open Voice #85 — Project Voice was a blast. These were some of our highlights and insights

Hi {{first_name}}, We spent most of last week in the US, so we’ll start with our highlights and then
Open Voice #85 — Project Voice was a blast. These were some of our highlights and insights
By Maarten & Sam • Issue #85 • View online
Hi {{first_name}},
We spent most of last week in the US, so we’ll start with our highlights and then the news to close it off.
So, we were at Project Voice… What is Project Voice you might wonder? It is one of the larger Voice events in the world. We both went to give talks and meet with the industry and the people in our community.
The event consisted of 5 themed days:
  1. Pre-event workshops
  2. All about Amazon’s Alexa
  3. All about Google’s Assistant (and a Voice Award show)
  4. All about Samsung’s Bixby
  5. Trade show meetup (which we won’t cover below)
We each put in a couple of highlights/insights for each day.

Pre-event workshops
Sam
At the pre-event workshop day, the afternoon was filled with interesting workshops. My favorite was by Mark Tucker, on speech markdown. It’s a friendlier/faster way to add markup to your dialog. For instance, change the speed or pitch of speech, add a pauze, etc. What became very clear during this workshop, is that there is fragmentation between the various voice assistants in supported features.
💡 With not all assistants supporting all features, I wondered if voices (like the colors in Google Assistant) will also support specific sunsets of SSML, and that we’ll see even more fragmentation in the future. Are they a bit like fonts, where they might (not) support ligatures, etc.
Maarten
I liked Mark Tuckers Speech Markdown workshop as well. Speech Markdown is THE approach to ensure conversation designs use SSML (the markup language in voice to get pauzes, intonation and other stuff so it doesn’t sound like a monotonous robot) easily across Alexa and other assistant platforms without needing to get into code.
I asked if he has combined codes yet, like CSS styles, to work more on whole sets of chats instead of word by word. Not yet Mark said, but he liked the idea. Check out how Speech Markdown works here
All about Amazon’s Alexa
Sam
The most insightful presentation on the Alexa day was on hearables - the wireless headphones that are taking the world by storm. The talk covered how we’ll see these tiny devices change over the next few years due to progress in the hardware (better chips, more features to monitor body metrics from the wearer) and software (the new Bluetooth standard opens up a lot of new possibilities). It’s really too much to cover here, so I’ll be working to get both speakers in an episode of the Spacerace in the future.
📈 Most interesting stat of the day was from Nick Schwab (founder of Invoked Apps) shared that he’s seeing about 15% of his Alexa traffic coming from devices with a screen. I take this stat serious, as Nick has millions of users per month over his many Voice applications. He shared the number in his talk on how his voice apps paid for his Tesla (also covered by this CNN interview we shared back when it aired on TV).
💡 Also, it seemed like most people in the audience have an Alexa device with a screen at home — Oh, and Nick was kind enough to give me an Echo Show 5. Lucky me for sitting front row when he asked who didn’t have an Alexa with a screen.
Maarten
Tom Hewiston of Voicelab.io, is an experienced Voice game developer (his would you rather game was most used in the first years of Alexa), he shared an updated version of his 10 hot takes on Voice. We mentioned him and his hot takes before.
What I liked about it that it is a grounded way of looking at Voice today based on experience. Listen to the podcast here.
Sam shared a cool slide about Voice skill squatting. The new opportunity / threat depending how you look at it. 
Skill squatting ...
Skill squatting ...
All about Google’s Assistant
Sam
At the Google day Maarten and I both had 2 presentations. In one of them we shared the story on Open Voice. In the other I talked about opportunities in Voice and Marketing. It was the first time I presented these concepts outside of closed sessions with clients, and I was keen to see how The Netherlands stack up against a mature market like the US. The response was good, and I’m looking forward to present this story more frequently in the next 6 months.
💡 Google sponsored the Award show and drinks that evening and shared a bit of their roadmap. It was nice to see an overview of what they have done so far, and hear commitment to further grow the platform and abilities.
Yep, that‘s me...
Yep, that‘s me...
Maarten
My favorite on day 2 was Cathy Pearl’s opening keynote. She is like the lead evangelist at Google and author of the foundational book Designing Voice User Interfaces. She looked back at early steps in Voice like Wildfire.
The challenges she sees in the Voice space are:
1- privacy and trust
2- Accessibility and inclusion
3- pushing the envelope (MLF: making it work better)
Especially the first one was important as she highlighted that now your data is not stored by default, you even have auto delete and mor eis happening locally on the dives instead of going back to the cloud.
My personal highlight was presenting Project Zilver where I shared the research results for the first time. Video here. The report will be published next week. .
Later that day I met the people of Constant Companion. They deliver professionally what we are found the Project Zilver audience are looking for. You will be hearing more of them soon 8-)
Lastly it was awesome to meet and hear Brian Roemmele. A prolific thought leader in the space. He coined teh term “Voice first”, first.
What was new for me was that he compared the stage of developed of the Voice industry to that of the days of the start of CPM and MS DOS. Because that was before Xerox Parc Labs share their mouse and window concepts which later “inspired” Steve Jobs and the Macintosh team. I will cheer on that a bit. I see many comparisons with emergence moments such as AOL/Compuserve, Blackberry/Nokia etc. 
Other take aways from him:
- the AirPod = the gateway into continuous voice assistant interfacing. Siri lacks too much currently for this to be fully realized.
- highlighting “hyperlocal” - the importance of moving away from the cloud 
- at some point various skills and apps are going to all interconnect with each other. 
- we are about to enter a Voice winter if we don’t fix discovery and monitization in Voice
- Voice is next abstraction layer in the age of AI like the app economy was in mobile 
- People will find the most convenient way to get to information: All gravitates to Voice
That’s Maarten in Project Zilver
That’s Maarten in Project Zilver
All about Samsung’s Bixby
Sam
It’s hard to pick the best talk for this day. I liked both Adam Cheyer (founder of Siri and now building on Samsung Bixby) and Brett Kinsella (the person behind Voicebot.ai). Adam shared a timeline of voice assistants since the 90s, with a lot of great insights and demo’s. Brett had the largest collection of stats and graphs I’ve seen someone present for some time. Still, I think I have to go with Adams talk, as it was during this talk, especially the demo he closed with, that I finally started to get Bixby.
💡Bixby is somewhat different from Alexa and Google Assistant, but I never really managed to get a grip on how it worked. Throughout the Samsung day it became clear why: I had been confused about the potential and what’s possible now. And now that I’ve seen it, I’m enthusiastic (and a bit scepticus – this will likely not launch any time soon in The Netherlands). I’ll be looking into Bixby more. I even submitted the podcast I make as capsule (a Bixby voice app).
Maarten
Here the opening keynote was my best for the day with Adam Cheyer, co founder of Siri and now also of Viv labs which powers the Samsung’s Bixby platform.
Like Cathy he starts with a look back. His inspiration for his Voice work that now exactly spans half his life started with the Apple Vision video called the Knowledge Navigator .
He mentioned General Magic. A pivotal company even though they didn’t make it. Both makers of iPhone and Android come from there. It’s crazy they sat two desks apart and are now responsible for devices that are in the hands of billions of people. More than half the world! Watch it!
I liked how he shared his goal: “To make the assistant as a paradigm as important as the Web and Mobile.”
And rated the state of Voice today:
- ASR get an A (Automatic Speech Recognition)
- 100k devevelopers in the industry. There 18 mln devs globally who could co Voice. Long ways to go.
- Ok conversation models and decent natural language.. work to be done
- Services rate OK.. they are simple. Not at the point that it replaces the web. It needs to be be expanded.
- Theirs party experiences are too fragmented. Discovery of skills and use cases is poor.
- Multimodel also has a way to go. 
- Poor cross-domain interaction.
- Poor personalization.
He showed he is the A level thought-leadership in the space. He is the visionary with the leadership. He and his Bixby team may move their Bixby Voice platform from his 3rd starting position up to 2 or even 1. He gave everyone in the room belief in that potential.
Again he impressed with how Bixby works and how easy it is to develop on their platform.
Did you know that Samsung sells more TVs per second than number 2-7 of the TV rankings together. And they all come with Bixby now.
My favorite slide that day was from Brett Kinsella. It shows the various Voice assistants globally . There is more than Alexa and Google assistant, we just havent checked them out!
Overall / other
Sam
Over all there were a few things I realized:
The Voice world seems big, but is actually pretty small — a significant of the familiar faces were present.
The community is amazing, not just in The Netherlands — this is something I had hoped to see, but clearly felt while present. Most people were very accessible.
Twitter is the main channel for the Voice community — I used to be active on Twitter, but became a lurker over time. I promised a few people to try and get back into the habit:)
While we (as the Netherlands) are ‘young’ in terms of voice, we’re not far behind most of the more mature markets — In some aspects we might even inspire some of the more mature markets.
I didn’t really adjust to the US time, but still managed to go on significantly less sleep then at home — I got a lot of energy from spending time with these likeminded people and will try to go again to future events.
Maarten
This conference was more about insights and community than business. No major brand here looking for a path into the new medium. The fact that all were “all about voice” for 4 days made it very special. Also so good that many luminaries like Adam Cheyer (siri and Viv founder), Jef Adams (who led the team that built Alexa) plus Cathy Pearl as well as many others were so accessible.
For me as a fan boy it was great, it also is my favorite moment in the emergence of a new medium. It all is small and cozy. Soon, when the numbers ramp up the game will be more transactional and business like. Good and fun in regards to the results but not as warm as it is now.
Lastly, here are the base slides with the Amazon, Google and Bixby installed base numbers (not active users!) plus the Voice bot total overview.
100s mln Alexa devices, 100k skills, >100k devs, 100%increase in usage
100s mln Alexa devices, 100k skills, >100k devs, 100%increase in usage
Android is key here
Android is key here
TVs here
TVs here
Now that we have big numbers, we’ll be switching attention to Monthly Active Users (according to Brett)
Now that we have big numbers, we’ll be switching attention to Monthly Active Users (according to Brett)
Now, the news...
Elon Musk
Teslas will soon talk to people if you want. This is real. https://t.co/8AJdERX5qa
4:20 AM - 12 Jan 2020
Brooke Hawkins
For anyone who missed my main stage talk at #ProjectVoice about the new Voice Ethics Committee: here is all the info you need! My slides with annotations and links to further reading can be viewed here: https://t.co/QvsDPHMvdv
5:07 PM - 17 Jan 2020
17 ridiculous Google Home hotwords that actually work
17 ridiculous Google Home hotwords that actually work
That’s it. If you made it all the way to the bottom: well done! And if you’re new, next week we’ll be back with the regular program for this newsletter. Have a great week 👋
Cheers,
— Maarten & Sam
Did you enjoy this issue?
Maarten & Sam

Open Voice weekly newsletter

If you don't want these updates anymore, please unsubscribe here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue