How Toyota is Leveraging the Power of Machine Assisted Cognition (MAC) to enhance collective well-being
By: Charlene Wu, Candice Hogan, Kate Sieck, and Matt Klenk
We make decisions every day, even every minute. And yet, with all that practice, we’re not very good at making decisions as a species. We overthink some things. We judge too quickly on others. We ignore facts altogether, or marshal the ones we prefer in order to bolster our choices. Our emotions can wreak havoc in pivotal moments, and can also be weirdly absent when we need them. Consider perennial choices that we often make: which health plan, which ballot measure, which investment, which car. We try our best to consider all factors — from upfront costs to the likelihood that our preferences may change before we have the opportunity to decide again.
Then, there are big decisions we face, both personally and as a society. For example, with a new financial goal, which investment path to choose, or what policies or solutions to pursue to combat climate change.
In the end, humans are human, and all that mental effort can often result in decision fatigue and less-than-optimal decisions — with negative consequences for us, for others, and for generations to come.
Is there an upside?
Yes. We can learn. We can change. We are highly adaptive.
Now, perhaps more than ever before, the world is facing consequential decisions that will impact our planet for generations to come. If we are to live in a world that fosters collective well-being, that nurtures and sustains us and future generations, we need better tools and frameworks for better decision-making. And that is the mission of the Machine Assisted Cognition (MAC) group at TRI.
At MAC, we seek to enable decisions that enhance collective well-being for people and the planet. To do this, we create new human-AI interventions and capabilities that enhance cognitive efficacy and expand emotional intelligence to augment human decision-making.
Who is MAC? And what makes us suited to take on this mission?
We start by focusing on understanding humans — our instincts, perceptions, beliefs, motivations, etc. This emphasis on human behavior is what makes MAC unique compared to other AI research teams. MAC is composed of world-class research scientists from behavioral science, machine learning, causal inference, and human-computer interaction. We’re psychologists, data modelers, neuroscientists, engineers, anthropologists, technologists, and philosophers bringing our expertise and experience from academia, government agencies, policy think tanks, technology companies, and corporate research to pioneer the new scientific field of Machine Assisted Cognition. Let’s break that down a bit:
M = Machine. This refers to the technology that we’re building or leveraging, such as sensors, models, software, etc.
A = Assisted. Our approach to AI is focused on augmentation, not replacement. We believe that the partnership between humans and machines can be synergistic and is key to our vision for collective well-being.
C = Cognition. The aim of our technology is to engage and extend human behavior, and the decisions (or lack thereof) that drive it. We want individuals to become more deeply engaged with data and information, to help them shift from an instinctive to a more deliberative decision-making process when appropriate, yet also help people know when to trust their instincts.
Together, we aim to understand and predict human behavior by employing a multi-method approach:
- Behavioral Insights: We conduct experiments, surveys, interviews, observational research, and diary studies to derive insights about decision-making.
- Data Modeling: We integrate existing behavioral science theory and/or our derived insights with machine learning approaches to build interpretable models of decision behavior.
- Technological Interventions: We develop software that bridges the last mile between algorithms and people, by partnering with the user through the use of novel interfaces or applications that deepen a user’s understanding of data through visualization or personalization to augment their decision-making process.
What do we actually do?
TRI has ambitious goals. Yes, we develop technologies for cars and robots. But those technologies only serve to fuel a bigger aim: build a world that supports collective well-being and happiness for all. Reaching these goals also requires research and AI in behavior change — understanding the “how” and “why” of decision making so we can support people as they adopt these new technologies. MAC focuses on this human side of the equation. That is, helping people act on their intentions by using AI tools to reduce cognitive biases, all in the service of helping people make more efficient, satisfying decisions with consideration for others and the future. For example, understanding consumer preferences to build awareness and help reduce carbon footprints.
At MAC, we’re starting with areas that are rooted in some of the most important challenges of our time :
- Building technology to inspire consumer adoption of existing and emergent carbon neutral technologies
- Evolving safety through adaptive vehicle interfaces that tailor coaching experiences to the driver, situation, and context to enable everyone to be better drivers and increase mobility for all
- Amplifying empathy to enable decision-making that fosters harmonious communities
Where are we going?
Ultimately, humans are at the center of decisions, yet those decisions ripple out to impact other people and, ultimately, our planet. As product designers and managers, we decide what products to develop, and as consumers, we decide what products to buy and how to use them. As voters, we vote on the ballot measures that will impact the lives of our community. As family members and neighbors, we choose how we connect to and care for each other. As we consider the existential challenges facing the world today — from climate change to global pandemics to increasing political tensions — one path forward is for all of us to better understand how our behaviors affect others, and to consider one another as we make decisions.
No human has the cognitive ability to imagine all the possible consequences, empathize with all those who would be affected, and weigh the risks and benefits of each option. We need assistance to achieve that goal. To that end, we believe in building MAC technology that makes this possible.
If this sounds incredibly hard, that’s because it is! We’re aware of the big challenge ahead of us, but we also recognize the huge need for this in our world and the positive impact we can bring about for science, business, and society.
To stay in the loop about our work, follow TRI’s LinkedIn here. If you’d like to learn more about our team, please connect with Charlene Wu, Director of the MAC team.
