Google I/O: The good, the bad and the AI
A few months ago, I attended a Women Who Code meetup where one of the girls at the event mentioned that she was getting a ticket to the Google I/O conference. Before that moment, I was under the impression that you had to be a blogger, tech guru, or some famous tech geek to be able to attend the Google I/O. Apparently, that isn’t the case: I found out that anyone can apply. Per the website, tickets are distributed randomly; however, looking at the diversity of attendees I doubt that’s a case, and I am glad it is not. Attendees were from all over the world. There was a huge number of women representing women in technology, there were meetups for different organizations and groups in tech such as LGBTQ, Black, Latino and many other communities. The conference run from Monday 7th- till Thursday 10th May. On Monday, the different organizations had their own meetups that served as a kind of pre-conference party. I was invited to the Women Techmakers event where I had a chance to meet and speak with inspirational women and hear about how they are making a huge impact in their own communities. This was followed by the main event: the I/O keynote and developer keynote.
I expected the main keynote presentation to be something exceptional. I had imagined fireworks in the background and people excitedly cheering; however, reality was little different. A few hours before it started, the attendees were lined up and asked to take their seats. It was very sunny, I was getting sunburnt, and to top it off, it didn’t smell too nice. But when Sundar walked onto the stage, I forgot all about all that, at least for some time. The keynote began with a big reveal of the numerous improvements to the Google Assistant. Google showcased six new voices, one of those belonging to John Legend (if Amazon adds Cardi B’s voice, I am switching to Alexa). It’ll use these voices to remind you to say “please” before you bark commands at Google Home devices. If you forget to say please, you’ll get a gentle reminder that you’re a jerk. For those that remember, Google Assistant will tell you what a good boy / girl you are. The biggest feature was the Google Duplex: a realistic-sounding AI capable of making calls with real people to set up appointments, make reservations, or do all the other tasks you’d rather not pick up the phone to do. It sounds like a real person and says things like ‘ummm’ which makes it more human. This is a feature I’ve been wanting for a long time. Do you know how annoying it is to pick up a phone, make a call, have a conversation and all that? Just kidding, but for introverts like me who would rather go to some website that supports booking than make a call, this new feature is amazing. Additional features include Multiple Actions, which allows you to ask two questions at once and another feature that allows users to continue the conversation without saying “Hey Google” a second time. Photos also got an AI upgrade: with just touch of a button, it can brighten, colorize, or add additional styling options with minimal human interaction. It can colorize old black and white images, strip color from the background of an image, convert photos to PDF documents or photoshop your face when you have bad face day (just kidding, the last one is a feature I would love to have). Another big announcement was the google maps update. Google tells walkers to begin heading south, and then semi-expects you to know which way that is. By using the camera, paired with GPS, and a new VPS — virtual positioning system — Google’s trying to end the song and dance that begins and ends with walkers turning in circles until the cone extending from their blue dot on the map is oriented in the right direction. They incorporated augmented reality by adding friendly creatures to lead the way. It’s nice having a travel partner (starts sobbing). The next big upgrade was the launch of Android P. As an iPhone user, that’s when I started snoozing and left for snacks (priorities), so if you want to know more about it just Google it. In all the sessions that I attended, AI was showcased in one way or another. Gmail was not an exception, also upgrading with its own AI element: Smart Compose that attempts to match what you’ve typed with words that could finish the sentence or phrase. It’s essentially the same tech that powers predictive text in some of the messaging applications.
The main keynote and developer keynote were followed by smaller sessions. For these sessions, attendees had to register in advance and even then, there was a huge line to get good seats, especially if the session had AI in the title. The sessions I registered for were about web, AR, VR, lot and AI. I learned a few tricks at the Chrome development tools session, about different JavaScript libraries such as guess.js, AR.js, three.js, hammer.js and a few others. AR and VR sessions demonstrated updates in plugin and libraries for web, which was the first time I saw VR and AR experiences in Web, which was as nice as it is in apps. Even though I loved all the demo projects they showcased, I am still not sure what the everyday use of augmented or virtual reality in web is. Yeah, it is nice to see how a table will look in my living room or a coffee machine (which was as big as a table — AR gives unrealistic standards to all the coffee machines out there), but what are the chances that a user is going to use AR for shopping or for cooking (yes Food Network has those features). In the future, I can see let’s say Google Glasses having AR and VR properties that would make the experience more seamless. Maybe at the next Google I/O? Another interesting session was about web performance using AI which is basically the idea to use lazy loading and pre-fetching by predicting what the user is going to click or navigate to. That technique would increase web performance and increase the overall web experience. One of my favorite sessions was Machine Learning for everyone, it was a great session by professors Fei Fei Li and Greg Corrado. This was a great session to understand where we stand when it comes to AI development. Should we be worried about AI taking over the world? Well apparently we are far from that. During the conference, Greg Corrado said:
“We finally got artificial neural networks to be able to recognize cats, and we’ve solved this holy grail problem of image recognition, but that’s only one small sliver, a tiny sliver, of what goes into something like intelligence. We haven’t even scratched the surface. So to me, it’s really just a leap too far to imagine that, having finally cracked pattern recognition after decades of trying, that we would then be on the verge of artificial general intelligence. “
Anyhow not to bore with all the session details, if you are interested you can see all the sessions here.
One the second day, after all the sessions were finished, we had a concert by Justice. It was a great concert except no one danced. I admire Google’s decision to give out lighting tubes that at least made people move their arms, but that was definitely the only concert I’ve ever been to where people refused to dance.
In general, being at Google I/O was an amazing experience, not only because I could experience all the latest new technology, but also because of the people. I had a chance to network with developers from all over the world, make friends, hear from the tech leaders and get inspired.
Lastly, I would like to thank Celtra for sponsoring my trip to the Google I/O conference.
