As a member of the SAS Data Ethics Practice, I was excited to collaborate with teams at the SAS Hackathon to learn more about their ideas for trustworthy AI. Artificial Intelligence has the potential to make a difference in the real world, and partnering with the hackathon teams was a great way for my team to showcase SAS' commitment to promoting responsible innovation.
This year’s hackathon teams developed many projects that caught my attention and left a lasting impression. As a data scientist and advocate of trustworthy AI, I was intrigued by several of the projects and the societal impact these solutions could have.
Trustworthy AI examples in the hackathon ranged from using satellite imagery to assess building damages caused by earthquakes to predicting patient inflow to intensive care units.
I was particularly drawn to a few projects that used key elements of trustworthy and ethical AI practices in their solutions, and I wanted to share more about these trustworthy AI examples here.
Generating synthetic data in health care
Health care data presents many challenges, including privacy restrictions and small sample sizes when looking at rare diseases. Researchers are looking to overcome these challenges by generating new data sets that can be used to improve health outcomes.
Syntho’s project showcased the power of AI-generated synthetic data in health care. This approach demonstrated the potential to use health data more efficiently without compromising privacy concerns. Syntho demonstrated how synthetic data can accelerate the analysis processes crucial for advancements in health care. This solution not only addresses privacy concerns but also paves the way for new possibilities in health care research, analysis and decision making. With privacy and security at the heart of this project, I was intrigued and wanted to learn more about the work this team aims to do going forward.
Creating a centralized flood control system
Jakarta, the economic center for Indonesia, is sinking. Many experts predict it could be underwater by 2050. Is there a way to solve this problem with AI, and could that solution help other cities affected by rising water levels?
The remarkable submission by Jakarta Water Resources Analytics (JaWaRA) impressed both the jury and participants. Using IoT sensors to analyze real-time data, this project offered transparent insights and actionable information to manage and mitigate flooding efficiently and effectively. What struck me about this solution was its emphasis on transparency, human-centricity and inclusivity. This project considered the needs and well-being of communities affected by flooding, making sure that the solution was designed with everyone’s best interest in mind. JaWaRA’s project could empower community members, government agencies and other relevant decision makers to take a collaborative approach to flood control.
Fighting neonatal sepsis
Sepsis in the neonatal unit can be life-threatening and hard to diagnose. Could using AI to monitor vitals and predict the likelihood of sepsis save lives of premature newborns?
Among several exceptional projects, Team IN-STEP emerged as the winner, and their project left a lasting impression on me. Their solution, titled and stylized as “savINg liveS fighTing nEonatal sePsis” showcased the integration of human expertise with AI-generated insights in medical interventions. The team approached the sensitive nature of medical device data with caution and privacy, which exemplifies transparent and explainable techniques. The solution aimed to help doctors fully comprehend and trust the outcomes of AI models, thereby ensuring responsible decision making in medical interventions. To me, Team IN-STEP’s solution embodied elements of human-centricity, transparency, privacy and security.
A pivotal moment in pursuit of trustworthy AI
These projects, along with several others at the SAS Hackathon, demonstrated the transformative power of trustworthy AI in solving real-world problems. It’s no secret that the SAS Hackathon annually serves as a testament to the power of collaboration and innovation. Bringing together brilliant minds from around the world, the 2023 event was no different. It united curious people to use analytics and open source tools in the SAS® Viya® platform to solve issues. As a judge closely involved in this year’s event, witnessing these projects reinforced my belief in the principles of trustworthy AI and the positive impact these solutions can have on society.
1 Comment
These are truly exciting examples around the utilization of trustworthy AI, particularly in healthcare settings. When will these use cases be available to share more broadly to our customers? Thank you!