WWDC vs. I/O vs. F8

Thanks to my sickness leave, finally, I have the time to watch this year’s WWDC from Apple, Google I/O, and Facebook F8.

The general feeling for each:

• Apple is a symphony of software and hardware which is unparalleled to anyone. It keeps the same vision, targeting to pro users and pushing the boundary of best quality graphics.
• Google emphasizes from Mobile first to AI first, and share the AI technology to everyone (with Tensorflow). Open to everyone and covering all areas, from medicine to education, Google has such a big aspiration.
• Facebook wants to turn the camera into the VR/AR platforms, bringing gaming and fun so we can share and socialize more in the virtual world. Full of interesting use cases but kind of shallow regarding the vision.

WWDC

Tim Cook is the one hosting the whole keynote, with the thread of numbers. 1.TV, 2.Watch, 3.Mac, 4. iOS, 5. iPad, 6. Music. Very clear product matrix.

From one platform to another, the hardware and software improvements are like the two sides of a coin, interweaving and supporting with each other. The OS has a new version, and the hardware also does their job to launch several new models. Both help the experience faster, smoother and more beautiful looking.

Four operating systems from Apple

It is all about going to the extremes, to pursue better sensational experience. I found it super sexy for them to talk about a lot of hardware parameters though I don’t understand most of them. But these numbers make them powerfully convincing.

For example, the new Mac display has 500 nits, 10-bit dithering, 43% brighter, and supporting one billion colors. One of the improvements for the new iPad is Display Refresh Rate. This rate is doubled into 120 HZ which enables more smooth and crispy scroll experience and only 20ms latency with the pen.

Sexy parameters

The speakers are Apple guys as well as the targeted users, “pro-users” who can make the best of the combination of software and hardware.

I/O

For Everyone

Google shows a sharp contrast with Apple. It mentions so many times: open to everyone, accessible and available for everyone.

• Android Go, the new entry-level mobile phone. Seven people per day in Brazil get to the Internet for the first time. Now Android has 2 billion monthly active users. To bring the next few billion users to Android, they need to think entirely different from Apple. They also launched the platform called “Building for Billions,” to help developers efficiently build apps for those large group of people. For example, Youtube has launched “Youtube Go” version, specifically designed for people concerning the data cost.

I was totally touched by the story. Now I have an unlimited data package. This story reminds me just 1 or 2 years ago I need to constantly check my data usage in case my phone is out of balance. It is easily forgotten how hard it is at the beginning and how much calculation you will do to save more data. We should look back and help others.

• Another story from a 17-year-old high school boy using Tensor flow. His family immigrated from Pakistan to Chicago. He was completely self-taught on machine learning. Now he is building products to help improve cancer diagnose accuracy, which has a huge impact on human being.

The technology empowers everyone.

• Google for Jobs: Google CEO says again “ it will work for everyone, all levels, and all areas.”

To help people find a better match job, again Google is addressing important social issues.

• Expedition: the Google platform for education now also adds AR features.

I can feel the excitement, and the Wow moments kids will experience when they have this technology. The whole world becomes more interactive and exciting to explore. They are more willingly to search and look close and study those hidden secrets.

As a mom, I feel so thankful that Google can promote technology to the education area.

AI first

The key thread of I/O is obvious, from Mobile first to AI first.

Mobile means identity, location, and payment. Mobile first means shaping how we rethink the internet. Now AI comes, time for us to rethink how all goes again.

This is the key sentence of I/O 2017.

• Language: We can have smart reply in Gmail. Voice becomes an important modality in Google products.

• Vision: Computers can understand the attributes of an image. Google Lens can recognize what the flower is; read the bar code and help connected with wifi; recognize with the restaurant front and give you reviews

Cloud TPU for training Google.ai will be used to:

• research: Auto ML, learning to learn
• build tools/ infrastructure
• applied ai: biology, medicine

I/O design

Compared to the dark setting of WWDC, I/O goes to outdoors with very bright color and three display screens add a vivid feeling.

There are quite a few interesting designs in the keynote presentation. Firstly, the countdown. The technique is not new, but still quite fun as a prelude.

Secondly, the intro video. “This is the story of an idea.” It reflects the struggles the developers in this industry are facing. How you may go through ups and downs and may face a bright new future with your idea? Th interesting part is how the blue character with a stick (the hero) and the yellow egg (the idea) turn into I/O logo.

Thirdly, this I/O logo, especially the green tilt line, leads to a consistent system of elements across the presentation.

Facebook F8

As I mentioned at the beginning of this post, I see a lot of detailed and interesting examples on how to use AR and make life more fun. But compared to the sexy Apple and the aspirational Google, Facebook’s scope is still how to grow their business.

Facebook 10 years’ roadmap

I really don’t see how profound it is to launch things like “Frame studio” where artists can create different frame styles so people are more willing to share photos on Facebook.

Or the Messenger has more games inside so people can compete, and more bots from different businesses. It is very similar to what WeChat did.

Since I rarely use Facebook products, I don’t have the emotional connection with these products. So let’s just look at the examples of AR from Facebook.

Mark Zuckerberg talks about three important use cases of AR.

• Display information: showing directions, messages
• Add digital objects: a digital chessboard, a TV so you don’t even need to buy a physical one
• Enhance existing objects: turn your house into a castle

Mark Zuckerberg only shows up at the first part of the keynote.

Facebook added camera feature to all its apps. Now they want to turn these cameras into a leading AR platform. He also introduced the building blocks for this AR platform:

1. Precise location: you can type a word and transfer it into 3D and place on the desk, or you can add animations as several fishes swim around your breakfast bowl.
2. 3D effects: turn a 2D photo into a 3D scene and you can freely change the lightening no matter turn it off or change the light positions; even add more effects to this room, fill it with waters or bouncing balls.
3. Object recognition: it can recognize what is on the table. You tap on different objects. It will show you relevant interaction options. For the mug, you can add some heat steam; for the plant, water it with some rain; for the wine, show you the detail information and reviews, even online shopping link.

Examples he showed which I am full of doubts for them:

• Nike NRC: make everyday thing funny, so people want to share

The funny thing is only worth sharing for the first few times. People will still get bored. And maybe we have already shared too much.

• AR games: turn the table into a defense board game to play together with the kids.

How to play with the kids? The father keeps holding the phone, and the kid needs to leave the screen and hit on the empty table to do interactions?

• Facebook office: not just painting, it can achieve the effects you can get in reality, e.g. constant waterfall
• Leave a note to your husband on the fridge

• Leave your drawings to the restaurant tables
• Leave a recommendation to a restaurant

CTO of Facebook explains more on how technology evolves and enable your mobile phone powerful enough to real time rendering all the AR effects.

Step 1: AI powerful enough

Computer vision is evolving. Now it can recognize where the person is, how’s the gesture and other objects on the photo.

Step 2: make it available on the phone

Style transfer on device

Computer vision on device

Some announced products that I find interesting:

1. AR Studio:

2. Facebook spaces

VR is naturally a social platform. Because it has the magic feel of presence.

Facebook did some experiments:

• People actually smile every single time when the avatar of themselves takes picture in VR
• Two avatars’ hands get stuck together. Both sides flush and want to confess to their partners.

Facebook Spaces is now available at the Oculus store. Three parts of its experience:

• You: create the avatar you feel comfortable with. Use AI to automatically create several versions based on the photos you had on Facebook. And you can tweak it, changing the eyes, mouth or hairs.
• Your friends: invite friends to this space. If they can’t use VR, now you will give them a Messenger video call.
• Your space: Here is like a magic canvas. You can explore and create your own spaces with the maximum of your imaginations so that you will have quality time with the friends. Go to outer space together, go to the theme park, play Snooker, safari or camping, whatever experience you can share together.