An Interview with Dr. Deborah Duong, Director for AI Development at Rejuve and Director of Network Analytics at SingularityNET and Dr. Ben Goertzel, founder of SingularityNET
Dr. Ben Goertzel
Photo from Dr. Ben Goertzel
Recently, Ben Goertzel, CEO of SingularityNET convened the “COVID-19 Summit” to bring veterans in AI and Data Science researchers with epidemiologists, front-line doctors, and policymakers to look at how we handled the situation so far and what are the expectations going forward.
One of the main themes from this Summit was the need for complex systems models such as agent-based models to inform policy. During this pandemic, at times, every policymaker around the world felt they were going into this pandemic without the information that they needed, even though we’ve dealt with other outbreaks such as SARS, MERS, etc..
The combined power of artificial intelligence and agent-based models in a Complex Adaptive System can give policymakers eyes, ears, and additional intelligence that can add transparency to the decision making process.
Photo from Deborah Duong
Due to the technical nature of this topic, Dr. Deborah Duong, Director for AI Development at Rejuve and Director of Network Analytics at SingularityNET, gave a talk to explain the power of agent-based models combined with artificial intelligence, its usage, and the information it can provide policymakers and practical professionals in the field.
To be prepared for the next pandemic or other types of disasters that may shake up core parts of our society, we will need a Complex Adaptive System.
Complex Adaptive Systems Can Give Us A Holistic Picture
A Complex Adaptive Systems, or a system that combines the power of artificial intelligence with agent-based simulations can fundamentally shift the way that we analyze data.
Dr. Duong says, “Complex Adaptive Systems are systems where the whole is greater than the sum of the parts. We can learn more about the parts from the whole. The parts also adapt to and change in reaction to the whole. You will have interactions between the micro and the macro.”
For instance, if you think about the Covid-19 pandemic, even before social distancing policies were in place, some citizens in some parts of the world started to wear face masks immediately as soon as the outbreak was reported. These people changed their behaviors in public immediately. They were changing their micro-interactions with the world around them. At the macro level, because of these people’s behavioral changes, in some parts of the world, it was easier for governments to contain the Covid-19 pandemic. It was easier get everyone else to follow social distancing policies, too. In this case, the micro and macro interactions, determined the outcome of social distancing policies.
Dr. Duong says, “Data alone doesn’t tell us much. But, data and patterns can inform policy. Complex Adaptive Systems can analyze spatial data patterns and conceptual data patterns to help to inform policy if we are careful of how to handle this data.”
Inspired by Dr. Michael Snyder’s work of collecting and measuring his own health data to analyze his body’s inflammatory responses, Dr. Duong and her teams use anomaly detection algorithms to analyze signals from wearable devices with an app developed by Rejuve to collect individual’s reactions during this pandemic. This data inspired her to modify the Complex Adaptive System that she developed with Dr. Ben Goertzel for SingularityNET to apply it specifically to the Covid-19 pandemic.
Dr. Duong says, “Healthcare workers and essential workers need to go to work during the Covid-19 pandemic. That puts them in danger even if they are wearing masks and gloves. They should be equipped with more information about their own health and the likelihood of infection in their places of work to make informed decisions. If they are empowered with a wearable that can alert them of the imminent infection of Covid-19 or if the probability of infection is extremely high, they can immediately decide to isolate themselves from their families, etc..”
Complex Adaptive Systems can potentially help us find a “Covid-19 data signature” from observations made from interaction data collected inside the population of people who are infected and who are not infected.
By discovering patterns using Artificial Intelligence and causal inference, conceptual groups can be identified and data can be analyzed in the context of what is happening within the society.
Data Sovereignty, Privacy, and Security
The media does a good job of scrutinizing AI systems for privacy, data ownership and security issues. It is possible to build a Complex Adaptive System that gives individuals their data ownership, preserves privacy and is secure. At the same time, through Artificial Intelligence with causal inference, a decision network can be created to inform policymakers. Much like the Markov Decision Process, data can be modeled in simulations. A percentage of the population can wear the wearables so that the least amount of data necessary for accurate decisioning can be collected and used for policy.
Dr. Duong says, “If you have a fully secure (encrypted) and private wearable where the AI is in the business of discovering patterns for decisioning instead of identifying individuals, then this device can inform people who are wearing it and empower them with more information to make their own decisions. At the same time, policymakers can obtain more intelligence from the decisions to base policy on.”
Nuance Will Allow For More Specific and Responsible Policy
Due to the widely used statistical methods, and the uncertainty surrounding these methods, policymakers during this pandemic didn’t have a lot to go on when encountering specific nuances in the way that people reacted to the pandemic.
At the same time, when social distancing or travel policies were set, there were specific circumstances related to certain groups of people where policies could’ve been set tailored to that group.
For instance, initially when policymakers recommended social distancing, there were questions such as what distance should people distance themselves? or should it be 3 feet or 6 feet?
At nursing homes, where healthcare workers battled the widespread of the virus, there were issues of healthcare workers transmitting the virus between two nursing homes because, in order to make ends meet, they had to work two jobs.
Dr. Duong says, “In a Complex Adaptive System, the system can adapt to these circumstances and be responsive to these nuances of changes. Not only can it recommend individual measures for people who are wearing the device, policymakers can see how these nuanced changes affect the entire system, our society. If we have a large population of nurses who are holding two jobs transmitting the virus between nursing homes, then a policy may be called for.”
The Greatest Benefit
One of the greatest benefits of using a Complex Adaptive System to analyze data from the Rejuve/COVID-19 app is that “flattening the curve actually means flattening the curve”. We have the misconception of “flattening the curve” during this pandemic. Everyone thinks that the curve will be flattened when following social distancing measures, that the number of people infected will shrink. However, that is not the case.
Photo From SingularityNET
The number of people who are infected and who could die will still be remain the same. But, it will simply take them longer to get infected and crowd our healthcare system.
On the other hand, if you use a Complex Adaptive System to set policy during the pandemic, you can lower the peak of the curve and truly “flatten the curve”.
Photo from SingularityNET
Dr. Ben Goertzel, “Complex Adaptive System is the start of fine-grained modeling. You don’t want Big Brother AI to collect your biometric data. You want high-level integration. You want data sovereignty. You want security. At the same time, policymakers need to understand the data, understand the impact and to be informed enough to set policy. Using an open-sourced Complex Adaptive System within the ecosystem of SingularityNET, we can achieve that. As we approach artificial general intelligence, the democratization of AI is crucial.”