AI can ‘help us move mountains’ for people and planet, Watson developer says

  • IBM Master Developer Neil Sahota believes artificial intelligence (AI) can help humanity ‘move mountains’ in terms of improving lives and the environment.
  • Sahota helped develop Watson, the supercomputer which is now being used in a variety of useful ways, like predicting crop yields for farmers in Africa.
  • In this interview with Mongabay, he shares multiple examples of AI being used by actors ranging from the UN to NASA and NGOs, for good.

Neil Sahota is an IBM Master Inventor and World Wide Business Development Leader in the company’s Watson Group. He works to create solutions powered by Watson, the supercomputer that he helped to develop which famously competed on the TV quiz show Jeopardy! against two human champions in 2011 and won. Sahota is a big believer in the power of artificial intelligence (AI) to improve the lives of people and the health of the environment, and while technology won’t solve all human or ecological problems, it has an important role to play, prompting Mongabay to ask him for an interview.

An Interview with Neil Sahota

Erik Hoffner for Mongabay: At our conservation tech site Wildtech, we increasingly publish news about how AI and machine learning can be applied to conservation, from using eBird to track songbird populations to applications that can curtail illegal rainforest logging. What are your favorite examples of how AI can aid the environment?

Neil Sahota: One of my personal passions in life is efficient water usage and conservation. In my pursuit of AI for Social Good, I’m always looking for opportunities in water management. One favorite example is Project Lucy, where IBM Watson is using AI to enhance African infrastructure, including water management in terms of dealing with water scarcity, and helping to maximize water use in agriculture. Imagine developing new farming techniques to grow more with less water! Another great example is Optimized System Controls of Aquifer Resources (OSCAR) by JEA where they’re boosting their water management capabilities by moving to “just-in-time” water. This is a great example of where they’re using AI to predict consumption patterns and prepare accordingly. In the future, they might be able to leverage AI and these consumption patterns to improve consumer behavior. Thus, they could help improve water conservation from two different points.

Project Lucy infographic courtesy of IBM

Mongabay: How has Watson begun to be used for social good?

Neil Sahota: We’ve started an across the board initiative called IBM Watson for Good. This spans across a lot of different domains like Project Lucy and Chatbots for Good which provide free help in some areas of law, therapy, and quitting smoking to name a few. In addition, the environment is a big focus for IBM Watson. We’re looking at solutions [to] better understand nature’s ecosystems. It’s a lot of ground to cover which is why we’ve opened up our technology for people to use, especially for social good. Hopefully by enabling people and organizations, we can build more sustainable and eco-friendly solutions much quicker.

Mongabay: I’ve also read that the cognitive computing of Watson’s ‘sister’ in Africa (“Lucy”) is being used for applications like improved waste collection in Nairobi. Can you say more?

Neil Sahota: Lucy is about possibility, and the work in Africa is showing how real these possibilities are. AI technology is opening up whole new opportunities that may allow us to improve things more dramatically than we realized. Take the waste collection in Nairobi for example. Improving route maps and adding sensors seem like obvious ideas today. However, Nairobi is the 4th most congested city in the world in terms of traffic and also suffers from poor conditions on some streets. By leveraging AI, the trucks can limit their exposure to potholes, bumps, and other obstacles to reach more locations for waste pickup, which has led to a reduction in the number of garbage heaps in the city. In turn, the trucks are helping the city to collect road condition information so that they can improve how they allocate infrastructure repairs. Likewise, Lucy is helping by using “telephone farms” (sensors that stream data on variables like soil moisture levels which are analyzed and sent back to farmers on mobile devices) in Kenya to collect crop data. With this type of infrastructure and cognitive computing power, Lucy is able to provide analytics and recommendations to the farmers on water usage, insect control, and yield forecasts.

Mongabay: How do the AI program development teams you’re associated with decide what data to collect, and how are the end users involved in determining the need?

Neil Sahota: AI is really about training. We have to establish the basic rules for decision making or “ground truth.” Think of it as teaching a child. We’re going to have our own perspectives on what’s trustworthy, important, and right. Thus, even if you have four organizations trying to improve crop yields using AI, they might have slightly different recommendations because each one will have a slightly different ground truth. As a result, the data (and access to data) is very important, but what’s even more important is the training strategy for the AI.

AI data processing helps researchers assess the health and conservation status of vulnerable koala populations.

Mongabay: You’re also working with the United Nations to develop a model and set of metrics to encourage nations and NGOs to pursue AI solutions for a more sustainable world. How exactly is the UN tapping AI for achieving things like the Sustainable Development Goals (SDGs)?

Neil Sahota: According to the U.N., there’s about a $US 5-7 trillion shortfall in investment to fulfill the SDGs. To bridge this gap, we’re looking at emerging technologies like AI to help. To entice governments and enterprises, the UN is developing a model to show the benefits of AI investment for initiatives aligned with the SDGs. The focus is twofold: 1) motivate more innovation, and 2) incentivize people to think about social good, in conjunction with commercialization. As a way of spurring adoption, we’re looking at a set of metrics that would give enterprises, agencies, and even countries an “SDG score” so they can gauge how successful their efforts have been so far.

Mongabay: And individual governments like Singapore are using this kind of tech for things like beach protection now, yes?

Neil Sahota: There are lots of examples of how artificial intelligence can help us protect our environment. Singapore’s beach protection is great example of how a government committed to environmental protection and which makes the investments can do a lot of good. Another great example is NASA’s Pre-Aerosol Clouds and Ocean Ecosystem (PACE) which is using machine learning to track the distribution of phytoplankton. While this may not sound like much, it actually has a profound impact on CO2 levels, and in turn, climate change and how plant and wildlife get impacted. Digital Crust is another great example of AI for environmental protection. It’s focused on a better understanding of subsurface processes in the planet which help understand impacts on groundwater and aquifers.

MODIS/Aqua true color image showing riverine outflow and phytoplankton blooming in the Gulf of Alaska, with clouds to the south and snow-covered land in the north over Alaska on April 12, 2017. Credit: NASA GSFC

Across these examples, AI is helping us to consume lots of real-time data and look at millions of variables and analyze millions of possible trade-offs and outcomes. As a result, we are developing a much better understanding of what’s happening to the environment, what could happen to environment, and what steps we can take to improve or avoid environmental issues.

Mongabay: What do you hope computational computing/AI/machine learning can be used for next, in terms of reshaping how people understand and use resources sustainably, to fight hunger, etc?

Neil Sahota: My hope is the AI will help us as people become what I call “sustainably integrated.” At a macro level, people generally understand the big challenges we face. At a micro level, it’s much more difficult to recognize the impacts. For example, I stopped shaving with water to reduce my water consumption. When people hear this, they often say it doesn’t move the needle. In the grand scheme of things, it probably doesn’t. On average, we use about 1 gallon of water to shave our faces.

However, what if my attitude inspired 100 people to do the same thing? Ok…. But does 100 gallons of water per day move the needle? What if those 100 people inspire 100 people each? Does 10,000 gallons per day move the needle? At some point, it does. This is where I hope we can leverage AI. If we had a little AI sustainability assistant to increase our awareness and help us change our behavior, even in a few small ways, it would really add up across the entire population. This won’t just move the needle for us; it’ll help us move mountains.

This interview has been edited for clarity.

FEEDBACK: Use this form to send a message to the author of this post.

This story first appeared on Mongabay

South Africa Today – Environment

This article is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.

You may republish this article, so long as you credit the authors and Mongabay, and do not change the text. Please include a link back to the original article.