Fun-fact: “I/O” stands for input/output, as well as the slogan “Innovation in the Open”.
“Smart compose” in GMail
This year, I actively started using “Smart reply” while using my Gmail mobile app. So, seeing a “Smart compose” seems like a natural progression of Call and Response. Only that instead of starting by first rolling out the “call” (which in this case would be the smart compose) they started with the response (smart reply). Perhaps, it was easier to figure out predictive replies than compose.
Whatever it is, this would be interesting if only it works accurately at least 50% of the time. Because I feel there are several permutations for the direction in which a sentence might be constructed. Given the current suggestive capabilities of the Google’s phone keyboard (G-board) I feel it would take a little while before they get sentence construction right. Thankfully, it is an intelligent system, so it LEARNS and would IMPROVE with use.
How-to use: Simply perform the same auto-complete gesture you use on your keyboard like hitting Tab.
Google’s Duplex
LOL, no, it is not a physical building. It is the name given to Google’s AI system for accomplishing “real-world” task over the phone. You can think of it as an AI-powered voice calling system. Listen to this conversation, can you guess who the human is?
If you were unable to guess, that means, the Duplex might have passed the Turing test.
As exciting as it might sound that your phone can take over the hassle of negotiating a reservation for you, it raises privacy and ethical concerns. Questions like;
How do I know who I am talking to?
What if one Duplex calls another Duplex to book a reservation without human intervention?
Spam calls?
However, Google was reported in a statement to have ascertained that there will be “disclosure built-in”. That way, one knows who is at the other end of a call.
We are designing this feature with disclosure built-in, and we’ll make sure the system is appropriately identified. What we showed at I/O was an early technology demo, and we look forward to incorporating feedback as we develop this into a product.
Improvements in the Google Photos ecosystem (Camera + Computer Vision + Cloud storage)
The third Google I/O announcement I am incredibly excited about is the improvement to the Photos ecosystem especially with regards to photo processing.
For instance, if you take the picture of a document, your Google Lens can recognise, convert the photo (.JPG) to PDF and save it for later. Likewise, it can help rejuvenate old (usually black and white) memories by brightening it with colour.
In addition, Google Photos ecosystem (Camera + Computer Vision) can help answer questions like “What’s in front of me?”. Also, they are experimenting with an AR navigation guide and soon they will be set to roll our a VPS, Visual Positioning system!
In Conclusion
Google announced a lot more builds like Slice API at their just concluded conference but I believe changes in products with 1 Billion Users like Gmail and over 500Million users like Photos should be more broadcasted.
Bonus: Popular American singer and songwriter, John Legend’s voice will be coming later as one of the additional voices for the Google Assistant.
Catch-up with all the action here in 14minutes.
