The use of AI to detect alcohol with Samatha Salim and Benjamin Riordan

In this episode, Dr Tsen Vei Lim talks to Samatha Salim, a PhD candidate, and Dr Benjamin Riordan, a research fellow, both at the Centre for Alcohol policy Research at La Trobe University, Australia. The interview covers Samatha and Benjamin’s article comparing the accuracy of artificial intelligence (AI) models to detect alcohol in video images.

• What are AI models? [01:18]
• The context of why this research was undertaken [02:40]
• Why is it important to detect alcohol in video images? [04:05]
• The regulations surrounding alcohol in media content [05:57]
• The three AI models used in the study [08:10]
• The key findings from the study [11:00]
• Are AI models better than humans in detecting alcohol in media content? [13:15]
• The implications of the findings for policy and practice [15:30]
• The key takeaways from the study [19:29]

About Tsen Vei Lim: Tsen Vei is an academic fellow supported by the Society for the Study of Addiction, currently based at the Department of Psychiatry at the University of Cambridge. His research integrates computational modelling, experimental psychology, and neuroimaging to understand the neuropsychological basis of addictive behaviours. He holds a PhD in Psychiatry from the University of Cambridge (UK) and a BSc in Psychology from the University of Bath (UK).

About Samatha Salim: Samatha is a PhD candidate at Centre for Alcohol Policy Research, La Trobe University in Australia, working at the intersection of artificial intelligence and public health. Her research focuses on quantifying alcohol exposure in films and digital media using scalable AI approaches, including multimodal large language models. By analysing large media datasets, she generates population-level evidence on the prevalence and patterns of alcohol portrayals and their potential influence on behaviour. Her work aims to bridge methodological innovation in AI with public health impact, supporting surveillance systems and informing policy interventions to reduce harmful alcohol exposure in media environments.

About Benjamin Riordan: Benjamin is a research fellow in the Centre for Alcohol Policy Research (CAPR). His research interests are broad, but predominantly, he focuses on using emerging and new technologies to understand and intervene with young adults who use alcohol. At CAPR, he co-leads the research stream on alcohol, media, and emerging technology, which focuses on understanding: 1) How is alcohol depicted or discussed in media (e.g., social media, films, music)? 2) What is the impact of exposure to alcohol-related content in the media? 3) What are the opportunities for policy change or interventions?

Original article: Comparing the accuracy of artificial intelligence models to detect alcohol in video images https://doi.org/10.1111/add.70337

The opinions expressed in this podcast reflect the views of the host and interviewees and do not necessarily represent the opinions or official positions of the SSA or Addiction journal.

The SSA does not endorse or guarantee the accuracy of the information in external sources or links and accepts no responsibility or liability for any consequences arising from the use of such information.

Music provided by Jack Shakespeare.

Hosted on Acast. See acast.com/privacy for more information.