Amazon’s Inspiration

For the past few days, I’ve been attending a conference called ReMars. One can say it is organized by Amazon, but what makes this conference different from the many I attend every year is that it is not all about Amazon. Jeff Bezos and his team have been organizing a conference called MARS for years now. It’s helpful to know what MARS stands for, which is Machine Learning, Automation, Robotics, and Space. These are essentially the core categories you can expect to learn something about by attending MARS. MARS has been for Amazon a source of inspiration and a hope to inspire others by having guests, who are generally acclaimed as the top scientists, academics, or entrepreneurs in their fields, to share ideas and some of the ground breaking work they are doing. MARS has always been only open to a very select group of people, but this year, Amazon did a very Amazony thing and scaled the conference.

While it was still invite only, the group expanded to several hundred but still had a goal of being intimate, communal, and bringing the best and brightest minds together to share ideas, challenges, and inspire one another. As I said, what made this event so different is how it was organized by Amazon but not about Amazon. The themes were still focused on Machine Learning, Automation, Robotics, and Space, and the best analogy I can make is ReMars is somewhat like TED but focused just on the MARS themes.

Even though this conference is not about Amazon, the same way that other company events I attend are all about their products and their announcements, Amazon did fit in their own learnings about Machine Learning, Automation, and Robotics and gave key executives some air time to share what they learned and what excites them about the problems they are setting out to solve. And yes a few announcements snuck in like Alexa getting a much more conversational interface and the official reveal of the Prime Air Drone and a go to market for Prime Air delivery to customers. in a few months. The rest of the talks taught us about how machine learning is being used in biochemistry to help solve health problems in bioscience. We learned about how far robotics has come and what major breakthroughs have led Boston Robotics to have drones that can walk and jump and run and scale objects very much like a human. We also learned the behavioral science around humans interacting with robots and the ways humans treat robots tell us very much about our humanity at its core.

I’ve appreciated the thought provoking sessions and wanted to share a few highlight takeaways.

• Humans and Robots living and working together. Until you experience how many robots Amazon has built and has running in their warehouses, you can’t appreciate what Amazon executives call a symphony of humans and robots working together. On the show floor, and at the conference, Amazon displayed the many robots they have designed and built to help them automate their warehouse work as much as possible. We heard a story about a factory in Japan where human workers show up to work and do some stretching and body warm up exercises and the robots they work with are programmed to do the same exercises. This has a single goal in mind, which is to help the workers feel more at peace and connected to their robot working companions. Since much of this particular collaboration between humans and robots is in warehouses and not in public, we don’t see this dynamic, but we are rapidly approaching this idea of humans and robots in an active community. So the question came up as to how we humans should think about these robots? Are they peers and colleagues, or essentially on the same level of humans or something else? MIT Researcher Kate Darling offered a profound observation and way we could think about robots. Through the years, she explained, humans have lived communally with animals in a working relationship as well as a companion relationship. So perhaps it is best if we perceive robots in a similar way we perceive animals. Fascinating, and worth a good think.
• AI May Really Help Solve Some of the World’s Greatest Problems. Yes, we hear this line, and it feels cliche at this point, but many of the worlds top minds in the field of AI truly believe this. We heard examples of how machines trained on computer vision to detect tumors were doing better jobs of predicting anomalies and specific treatments than expert physicians. Or how these machines could predict with greater accuracy the severity of an injury. We learned how the cost of manufacturing a new drug for an illness has gone from $100m to over $2.5 due to many more failures in the trial and error process to end up with a winning compound. AI seems well positioned to run simulations of these compounds and help bioscientists narrow the field of potential compounds and then test to see their effect. I’ve believed AI will be one of the, if not the most, transformational technology many of us will ever witness in our lifetime, and I believe it even more now.
• One Observation about Amazon. Yes, this was not a conference about Amazon, but there is an interesting Amazon observation to be made. In the keynotes, Amazon employees gave us we heard about Amazon’s robotics strategy and what they learned, solving challenges in automation with robotics. We learned about how Amazon’s AI models are helping to make shopping on Amazon, or using Amazon services more relevant and personal and provide a better customer experience. We learned about how Amazon created their Just Walk Out retail technology showcased at Amazon Go stores and more. And what hit me, was that while Amazon wasn’t here to pitch AWS to the world, Amazon, as a company is the first best customer of AWS. With this viewpoint, Amazon has then built AWS on the back of the learnings of a company as good as scaling technology as anyone and across many industry disciplines. These learnings, and the solutions their learnings led to invent, put them in a position to then offer these solutions to hard problems as a part of AWS to other customers. AWS was built out of Amazon needing a solution to solve their problems and then became a platform to help other people solve theirs. Now those tools included in AWS include deep expertise in machine learning, computer vision, automation, and more. I’ve long felt competing with Amazon in key areas like retail and commerce will be very difficult, and I believe that even more now.

At ReMARS, Amazon hopes that invitees are inspired by the work being done by the speakers at the sessions but also that it inspires them as well to keep charging forward and inventing the future.