Through June 2023, we will introduce interdisciplinary artificial intelligence (AI) research at University A. This shows how universities are leading the way with the aim of making AI safer, more reliable, and fairer.
Nidhi Hegde is an Associate Professor in the Department of Computing Sciences at University A, a Fellow of the Alberta Institute for Machine Intelligence (Amii) and Chair of the Canadian CIFAR AI Committee.
In this week’s spotlight, Nidhi discusses how ensuring fair results across different racial, ethnic, gender and age groups in machine learning models contributes to the responsible development of AI. increase.
What is AI?
People use the term AI to refer to algorithms designed for specific tasks that learn from and ultimately act on data. AI is really an umbrella term and what I just described is more like machine learning. AI includes machine learning, but also includes things like robotics and computer vision. In general, automatic adaptation to data, or learning how things should work, is called AI.
Briefly describe your research area and how AI is involved in it.
My research focuses on privacy and fairness in machine learning. I am interested in whether a trained machine learning model is privacy-preserving in the sense that it cannot or cannot infer personal information about individuals not already present in the data. A person was involved in training a machine learning model.
Fairness refers to the results of machine learning models and whether these results are similar across different subgroups within a population. We can more precisely identify major and minority subgroups along race, ethnicity, gender, and age—many of what we consider sensitive traits and demographic factors. I don’t want to know. In that sense, we don’t want the output of machine learning to be unfair. I’m interested in analyzing whether there are unfairness issues in machine learning models, and more importantly, designing algorithms that are already fair with respect to these subgroups.
How is AI impacting our lives and what are the common misconceptions people have?
AI is impacting every aspect of our lives in many ways. It may be the little things we’ve grown accustomed to for a long time, like Amazon recommending you a book, but the new bus route introduced because the data was used to determine it’s best It could also be In healthcare, AI-powered diagnostic tools may be developed. I think it’s hard to find an aspect of our lives today that isn’t affected by AI.
A common misconception is what people expect from AI. A good example is the recent ChatGPT phenomenon. People were amazed at how this tool generated these texts and equated this with artificial general intelligence. Even though AI is trained to perform specific tasks, people expect too much intelligence or power from AI.
A common misconception I see in my work is that people think everything about AI is great. It simply makes our lives easier or automates everything. Because I approach my work through a lens, I can see that there are other AI impacts that we don’t pay enough attention to. Keep in mind that these AI tools and services are built for a specific purpose. We’re not quite there yet with robots locking our doors, but we still need to be aware and aware.
What does the long-term future of AI look like, and how is the U of A leading in this space?
I think the answer will always change. Things that seemed distant a few months ago suddenly appear in a short period of time. So whatever happens with AI, we have to adapt. Eight months ago, we had no idea that so many people would be using ChatGPT and how it would affect so many lives in so many ways. We couldn’t have imagined this multimodal model that could generate text based on all these different kinds of data, or a diffusion model that would create an image.
What I would like to see in the long term is responsible AI development that will have a positive impact on us, and concerted efforts for collaboration, where AI is developed evenly across different types of research institutes. It is to Large companies that can afford the resources they need for all their AI work.
In the long term, AI will be practically embedded in many parts of our lives. A I think the university is in a very good position for that. We not only have a very strong computing science department that studies the core algorithmic nature of AI, but we also have a number of other faculties and departments that are responsible for the engineering part that applies AI to various fields. . You will see great growth in this. For decades, people have been doing the basic work that makes algorithms work and scales AI at scale. We are now at a stage where we can implement various aspects of this, and we can already see different university departments using AI in their respective fields. There is a lot of strong research going on at the university.
Over 20 years ago, the Government of Alberta invested in Amii, the promise of AI here in Alberta. Since then, continuous investments have been made to grow the local ecosystem through groups such as CIFAR and Alberta Innovates. This support has helped create the conditions for industry, academia and government to work together to accelerate positive impacts on society. This is a time of great opportunity for the AI space.
What do future job seekers need to know about AI?
People will likely not be in the same jobs in 30 years, so they will have to learn how to adapt to how AI changes their field. At the very least, we should be aware of the various uses of AI, such as data manipulation, data science, and machine learning. Techniques that lead to AI — because it impacts their work. They will have to adapt to new roles unique to humans, as new ways of automating things and making things easier will emerge.
This conversation has been edited for brevity and clarity.
Nidhi is one of 26 faculty members and Amii Fellows who are Canadian CIFAR AI Chairs.
Through new investments, Universidad Amy and Universidad Amii will soon have 20 faculty members whose work reflects the transformative impact of technology on health, energy, indigenous leadership and more. Learn how these new researchers are helping shape the evolving landscape of AI.
About Nidhi
Nidhi Hegde is an Associate Professor in the Department of Computing Sciences at the University of Alberta, an Amii Fellow and Canadian CIFAR AI Chair. She spent many years in industry studies addressing many interesting issues until she joined the Australian University in February 2020. Until recently, Nidhi was a research team leader at Borealis AI, the research institute of the Royal Bank of Canada, where her team worked on privacy-preserving techniques for banking machine learning models and other applications. . Prior to that, Nidhi worked for many years at research institutes such as Bell Labs, Technicolor and Orange.
The Innovator Spotlight is a series that showcases faculty with the discoveries, knowledge and ideas that drive innovation.
Do you know anyone at your university who turns ideas into amazing realities? It could be you! We are interested in hearing the voices of those who are helping to shape the future, improve quality of life, drive economic growth and diversification, and serve the public. It features people working in everything from accelerating solutions in the energy sector, shaping the evolving landscape of artificial intelligence, to carving new paths in health and indigenous leadership.
Please contact us at blog@ualberta.ca.
