Back when I worked at Gigaom, every year we’d invite Amazon CTO Werner Vogels on stage at our big cloud computing conference and interview him about the future of the Internet and distributed computing.
Nowadays Vogels is conducting his own interviews, heading out to far-flung locales to talk to innovators about how they are using cloud computing and AI to build products and change the world in ways both big and small.
It’s all for his new Amazon web show titled Now Go Build with Werner Vogels, and for his most recent interview, the cloud computing pioneer went to Singapore to talk to the inventors of the Rotimatic about why they built machine learning into a home food robot.
The video is a fun watch as Vogels visits with the wife and husband team of Pranoti Nagarkar and Rishi Israni in the middle of a bustling Singapore restaurant and later heads into the kitchen to try his hand at making his own rotis.
The conversation high point for me was hearing Nagarkar and Israni talk about why they decided to power what was, at first, a fairly simple automated roti maker with machine learning.
“I remember there is one particular phase where we tried to make it an embedded system where you have a fixed program that could run it and it could find out the right proportion of flour and water,” said Nagarkar, who came up for the idea of the Rotimatic after getting tired of making roti every day by hand.
“That wasn’t enough,” she said. ” You had to build machine learning in because every time you make a dough ball you learn something more about it and you use it for the next dough ball. It just kept making the machine better and better.”
Israni agreed and said they also realized machine learning would be necessary to ensure high performance of the robot over its lifetime.
“We see machine learning mostly in computer systems where things aren’t moving much” said Israni. “Here with life, there is degradation in all the tolerances and all the parts, so the performance of parts keep shifting over time.
Eventually, the conversation turned towards the challenges of today’s food system.
“I’ve realized a few big problems with our world,” said Israni. “Seventy percent of illnesses are lifestyle disease-related, and they are primarily related to the type of food you consume.”
The other big problem is food waste. “One-third of world’s food is wasted,” said Israni.
Israni sees these two challenges are intertwined. “On one hand, you have people who are unhealthy who are maybe eating a little more than they need and the other you have people who are dying because of lack of food. We find that these two problems are prime problems to be solved by the kitchen of the future.”
How does a more technology forward kitchen helping solve these challenges? According to Israni, through shared data and more connected appliances.
“This is an information data problem,” he said. “We already know how to fix the knowledge gaps, but we can also execution gaps by building machines that are fully connected and exchange information with one another to delivery a cooking experience.”
Vogels agreed.
We also need “to see all of these devices as a platform and not only as a single function device,” said Vogels. “Software eats the world.”
“But you can’t eat software,” replied Israni with a smile.
You can watch the full video below: