Jan 2, 2019 · 28m · a16z
a16z Podcast | QR. AR. VR.
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the a16z Podcast, Andreessen Horowitz partners Kyle Russell and Connie Chan analyze major announcements from Facebook's F8 conference, examining parametric QR codes, brain-computer interfaces, augmented reality, social virtual reality, and messaging chatbots through the lens of user adoption and cross-market tech strategy.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 6.8 Guest teaching 6.2 Guest disagreement 2.2 The host pushing back 4.2
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 26:22 Connie rejects discovery premise
Connie directly refutes Kyle's premise that discovery caused chatbot failure, asserting that lack of integrated payment credentials and real utility are the true blockers.
Hardest push from the host ▶ 16:29 Kyle counters filter fatigue with happiness theater theoryKyle explicitly offers a counterpoint to Connie's concern about AR filter fatigue, arguing that creative filters actually combat toxic happiness theater on social feeds.
Biggest teaching moment ▶ 3:09 Connie explains WeChat QR adoption and mini programsConnie educates Kyle on how WeChat's deliberate username-free architecture drove QR code adoption in China and introduces the power of mini programs.
The host holds their own ▶ 8:53 Kyle defends BCI utility with sentiment telemetryKyle counters Connie's practical doubts about BCI utility by articulating how non-verbal brain telemetry solves public audio privacy and captures implicit user sentiment.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Parametric QR Codes and WeChat Mini Programs | 5 | 7 | 1 | 2 | Kyle introduces the framing around US QR code adoption and draws parallels to Snapchat and Google instant apps, but invites Connie to explain WeChat's ecosystem. Connie educates Kyle on WeChat's account architecture, QR code friending mechanics, and the underappreciated utility of mini programs. | |
| Brain-Computer Interfaces and Emotion Tracking | 8 | 5 | 2 | 5 | Kyle displays deep technical domain expertise by providing a detailed history of BCI devices, gaming inputs, and Elon Musk's Neuralink. When Connie expresses skepticism about practical BCI typing speeds, Kyle pushes back by framing brain feeds as non-verbal privacy solutions in public settings and implicit sentiment telemetry for social platforms. | |
| Augmented Reality Filters and Social Media Dynamics | 8 | 6 | 2 | 7 | Kyle presents a structured overview of AR platforms, headsets, and social mechanics, explicitly pushing back on Connie's filter fatigue concern with his theory of happiness theater. Connie schools Kyle on WeChat Moments' private social record design, explaining how hiding likes and comments reduces social burden. | |
| Social Virtual Reality and Digital Identity in Facebook Spaces | 7 | 5 | 1 | 2 | Kyle and Connie engage in a highly collaborative discussion about Facebook Spaces, VR sociality, and digital identity. Kyle demonstrates expertise on avatar machine learning generation and transgender identity expression, while Connie brings up World of Warcraft identity dynamics. | |
| The State of Messaging Chatbots and Frictionless Payments | 6 | 8 | 5 | 5 | Connie directly rejects Kyle's premise that poor bot discovery is the primary reason for weak developer traction. She schools Kyle by reframing the core issue around missing payment friction reduction and lack of true utility, leading Kyle to concede the humorous disconnect between futuristic BCI and broken basic payments. |
Statements from this episode (13)
Chan: Most Americans do not know how to scan QR codes
“If you ask the average American how to even scan a QR code, even now, they probably have no idea.”
Chan: WeChat drove Chinese QR adoption by making it primary for friends
“In China, what really helped it take off is WeChat made it front and center, a primary way of adding a friend. So as opposed to having to scan for a particular username, because a username is not even required when you set up a WeChat account or knowing their …”
Chan: WeChat Mini Programs are underappreciated and their potential is unrealized
“Mini programs is something that they've launched, which I actually think is totally underappreciated. And you haven't really began to see, you haven't begun to see the power of the mini programs concept in China yet.”
Russell: Current BCI tech will allow seamless thought-controlled computer actions
“And now with this current generation, it seems like between Elon, Facebook, probably something from Google that We're actually going to start to see these things that kind of invisibly actually just let you think something and have some kind of action show up …”
Russell: Facebook brain-computer interfaces will track emotional reactions over typing speed
“What wouldn't shock me, despite the lofty goals of getting this hundred words per minute, You know, kind of success rate. I have a strong feeling that what we're actually going to see is kind of getting ground truth reactions to, for instance, posts on social …”
Russell: Microsoft HoloLens costs $3,000 and has only a few hundred users
“You look at HoloLens, and it's 3000 dollars, and maybe a couple hundred people own it.”
Russell: Public metrics turn Instagram and Facebook feeds into happiness theater
“So a little bit of a counterpoint or pushback against that is, you know, if you look at Instagram or Facebook, the nature of having metrics publicly exposed to people and comment counts and things like that means that content typically can devolve into somethi…”
Chan: WeChat Moments' private engagement metrics encourage freer social sharing
“So WeChat moments doesn't have that public count of likes or comments. And therefore, at least I look at myself and anecdotally with my friends, people share in a very different mindset. So if you post something on WeChat and say a mutual friend of ours commen…”
Russell: A socially acceptable Apple AR headset is still years away
“You have to make a headset that you're not embarrassed to wear in public, so maybe Apple can do it but probably gonna be a couple years.”
Russell: Screen usage perceived as antisocial is actually highly social behavior
“What looks like antisocial behavior is actually social behavior because there's more interesting people on the other side of that screen.”
Russell: Facebook trained ML models on celebrity photos to generate avatars
“Facebook made the system that they made a giant data set of celebrity photos plus bitmojis associated with those faces to train a machine learning model that then generalized to Regular people, and so it looks at your profile photo, then makes a kind of cartoo…”
Russell: Facebook was a poor steward of Messenger's early developer platform
“My kind of read on this has been that Facebook hasn't been, let's say, the best steward of Messenger as a platform, for instance, for the first six to nine months of it, other than the kind of Bots that Facebook itself specifically highlighted and linked to in…”
Chan: Chatbot adoption failed due to missing use cases, not discovery
“My concern is that the issue around bot usage and adoption is not actually the discovery piece. It's that the great use cases for the bots haven't been created yet. And that's more a function of Facebook educating developers or developers coming up with the ri…”