Seminar Recording
In this Data Science Under the Hood webinar, Dr Aloha Ambe presented, “What We Don’t Talk About When We Talk About Machine Learning”.
About this presentation
With Machine Learning (ML) systems on the rise, ML design practices are having more impact on more humans than ever before. To improve these practices and increase the benefits of ML to humanity, we believe it is fundamental to understand the forces and concerns that shape and motivate the practices of ML system designers – a heady mix of technical and very human issues.
In this presentation, I want to reveal insights into some of the forces that influence designers and the design process, but about which people rarely speak directly. My aim is to convince you (1) that to understand and improve ML design it’s important for us to go out “into the wild” and engage with real people, (2) that there are some truly profound issues that people don’t speak about directly in ML work, and (3) that some of these issues can be brought out in the open by asking:
- How close are you to the human consequences of the system you are designing?
- Who decides why and when your system is good enough to proceed? How?
- How well aligned are your personal goals to the overall desired outcome?
These themes (questions) are valuable because they address profound (but tacit) influences on ML design practice, opening a challenging yet necessary way forward in the design of ML systems. This presentation aims to offer a different perspective on ML design practice and aims to help develop new strategies to improve the process and outcomes of machine learning innovations.
About the presenter
Aloha Ambe is an early career academic in the field of Human-Computer Interaction (HCI) with a passion for understanding and amplifying the role of people in technology design. She completed her degree in Doctor of Philosophy at Queensland University of Technology (QUT) under the Design Participation Lab of the School of Computer Science. Aloha is currently a postdoctoral research fellow in the Data-Focused Decision Making Program, Centre for Data Science. She is part of a project that investigates the design of new systems in which humans and machines interactively learn and explore together to have a more reciprocal and synergistic relationship. With a human-centred perspective, she is working with machine learning and Artificial Intelligence experts (ML/AI) as well as domain experts who analyse, interpret and make decisions with data. Both ML/AI experts and domain experts play a critical role that involves workflows and processes that need to be understood, supported, and learned from so to improve people’s relationship with data that is positive and proactive.
In her PhD, she investigated older people’s perspectives, exploring how they might reimagine monitoring and tracking technologies in their life and future, by involving them in collaborative design (co-design) and qualitative research. Monitoring technology solutions for older people’s independent living tends to treat them as passive recipients of technology to be observed by others. Together with her supervisors, she developed design approaches that shift the emphasis from perceived needs to bring to light the lived values, agency, and aspirations of older people.
Details:
| Location: | Online |
| Start Date: | 25/02/2022 [add to calendar] |
| Start Time: | 12:00 PM |
| End Date: | 25/02/2022 |
| End Time: | 1:00 PM |
| Register: |