I spent a week at Google AI - here are the lessons I learned
I had several powerful takeaways from during my time with people from different teams at Google and there are lessons that you can start applying as well.
Earlier this month during Google I/O, Jeff Dean, Head of Google AI, announced the winners of the Google AI Impact Challenge which aims to provide coaching from Google’s AI experts, Google.org grant funding from a $25 million pool, and credits and consulting from Google Cloud. Out of 2,600+ applications, 20 projects were selected and the project that I have been working on at Wadhwani AI is one of them, receiving a grant of $2 million and being the only team from India to be selected in the Impact Challenge. As a follow-up, we had a booth at I/O where we had the chance to demo our solution and a team of 3 was invited from each grantee to attend a week long accelerator program to kick-off a 6-month mentorship period for each grantee.
Last week I attended the accelerator program in the beautiful city of San Francisco, alongside two of my colleagues, representing Wadhwani AI, along with the other 19 grantees belonging to different parts of the world. We were hosted by extremely passionate and energetic folks from Google Launchpad, Google.org and Google AI. The week was packed with 1:1 mentorship sessions, keynotes on various aspects of an AI solution, introduction to the design sprint process along with other best practices from Google and a whole lot of networking activities. The biggest takeaway from the week is that we have a concrete plan for what we intend to achieve over the next 6 months and an incredible support team in the form of a Grantee Success Manager (GSM) and an AI coach, who’ll work with us over the next few months and help us achieve the goals that we have set for ourselves, pooling in resources from various team at Google, as and when required. I’ll structure the post in two parts:
At the bootcamp: Here I’ll briefly go through what each of the activities entailed and add relevant links wherever possible for you to checkout yourself.
Learnings: This is where I’ll actually nail down my observations and key lessons that I intend to share.
Discover Latest Jobs in AI, ML, Big Data & Computer Vision:
At the bootcamp
As mentioned before, there were a bunch of things that happened during the week and rather than going into too much detail, I’ll try to highlight what each component was about.
On each of the first 3 days, we had a keynote session by someone from Google AI. One of them was taken by Jason Mayes, who works as a Senior Creative Engineer at Google (and our AI coach, YAY!) who talked about the creative applications of AI, like Google’s thing translator which translates a text in an image and rewrites the translated text back onto the image as shown below. The complete slide deck can be found here.
Another keynote was taken by one of Google’s all-time legend and the author of one of the most famous books written in AI, Peter Norvig. He essentially talked about the fact that most people obsess about the AI part, while building an AI product when actually the AI part is only about 10% of the entire pipeline, which consists of data collection, preparation, annotation, cleaning, analysis, engineering, etc. He made an interesting analogy that one can use backpropagation to tackle errors in the AI part, but there is no such algorithm to handle errors in any other component of the pipeline and hence, we should really focus on getting that right.
Finally, Andrew Zaldivar gave a talk on the ethics side of building an AI solution. Specifically, he referred to a large body of work initiated in Google through the People and AI Research (PAIR) team and I’ll mention two specific tools that he suggested:
People + AI Guidebook that provides a framework for designing human-centered AI products:
2. The What-If tool: This tool can be used to inspect the performance of any machine learning model including several features, some of which are shown below, and all you need to use it is Jupyter. Isn’t that amazing?
This was one of the most helpful components of the accelerator program. For two complete days, each grantee had several hour long sessions with various Google mentors, catering to the need of each grantee.
For example, our first day was full of technical mentors which helped us clarify various doubts regarding the algorithmic pipeline, the overall infrastructure and several best practices when it comes to building a scalable solution. However, we also had various product-based doubts and because of the diversity of the mentors, our second day was structured to attend towards those needs like Product Strategy, Fund Raising, Long-term sustainability, User Centric Design etc.
The idea of the Design Sprint was invented by Jake Knapp & John Zeratsky during their time at Google Ventures. The Design Sprint essentially provides a framework to test and validate ideas in a 5-day cycle by bringing together all the team members together in a room to focus on just one thing by dividing the sprint into 6 stages:
They wanted us to actually have the feel of being in a design sprint but since we had only 3 hours, only a few components among the ones mentioned above were taken on by dividing all the grantees into teams of 5 along with a common problem that each team needs to work towards solving.
On top of actually being in the process of a design sprint, it also turned into a networking session because of the fact that members from different teams were asked to come together and collaborate. The entire activity involved sketching, brainstorming, conducting interviews, writing down ideas on Post-Its and finally prioritizing the ones that received the majority vote.
One key rule of thumb that they stressed was to get EVERYONE on the team together — designers, engineers, product managers, researchers, even accountants if they happen to be on the team — as each person can provide a unique perspective unbeknownst to everyone else who might be wired to think only in a certain way. More details on the various components of the sprint can be found here.
Objectives & Key Results (OKRs)
OKR is the framework that Google uses internally to track project-level and sometimes, individual-level progress over a period of time. OKRs was a key focus of the accelerator program and all the grantees were supposed to set their respective OKRs for the next 6 months of mentorship. We were first given a talk by Zachary Ross on what OKRs stand for and how to set them correctly, along with examples of both good and bad OKRs. For detailed information on OKRs, refer to this guide by Google. Both, our AI coach and our GSM, were closely involved in the process of setting OKRs and we’ll keep checking in with them going forward on our progress on these OKRs.
Here are the key lessons that I take back from my experience and would like to share with everyone:
Punctuality: Being extremely mindful during meetings and ending them within the slotted time made sure that our discussions were focused and didn’t waver off to a million topics.
Regular acknowledgement: More often than not, we tend to focus on the “things that need to be done” and forget to celebrate “things that have been done”. An almost daily acknowledgement of the time and effort put in by various people from Google, be it the AI coaches or the support staff responsible for arranging our chairs everyday, was the culture that was set from day 1 and that translated to people who were really happy in what they did. As for the grantees, we didn’t see different teams like Launchpad or Google.org or Google AI. We only saw one team — Google.
The little things always matter: A strange thing that got everyone’s attention during the bootcamp was — the water bottle.
The design of the bottle given to each grantee team member resonated the principles that Google has always stood by — simple, easy to the eye and the best in what it does. This really has been the best bottle I’ve ever used and the first one I’m obsessing so much about. But other small things like goodies, food, even the water, were very carefully thought of, and for a company as big as Google to be mindful about these little things, is a big reminder for everyone.
Regular feedback, constructive criticism and lifelong learning: A major event at the end of each day was a call for feedback on the entire day. To put things into perspective, Google Launchpad has already coached 500+ startups till date and the fact that they are still looking for feedback is a good reminder that you can always improve. Also, after one of the sessions, someone that I had interacted with during the session, called me up and asked me if I would like to hear some feedback about the way I was asking questions. It is rare that someone would be willing to call you up and give you constructive criticism but it was a wake-up call for me that even if no one around me is doing it, I should not hold back.
The past week was one of the best weeks that I’ve had and adding in the fact that it was the first time that I travelled outside India, I witnessed a very different culture and met with people of varying backgrounds. From being inserted into an unknown environment to making new friends and exploring new places, it was a week that I will always remember. It was a week that I really needed.
If there’s anything that you might want to share with me or give any sort of feedback on my writing/thoughts, I would love to hear it from you. Feel free to connect with me on Twitter, LinkedIn, or follow me on Github. To stay updated with my posts, do follow me on Medium.
Don’t forget to give us your 👏 !
I spent a week at Google AI - here are the lessons I learned was originally published in Becoming Human: Artificial Intelligence Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.