Marco Zenone

@marcozenone.bsky.social

Assistant Professor @uOttawa researching medical & public health mis/disinformation, digital platforms, and commercial determinants of health. Zenonelab.ca

New paper out! 📢 The #wellness industry is a multi-trillion $ force: selling products, shaping norms, and now influencing policy. But it's rarely analysed in the way we scrutinize tobacco or pharma. We apply a commercial determinants of health lens to wellness, including #MAHA. Short thread 🧵⤵️

AI “safety measures” leave much to be desired. The authors found that chatbots refusing disinformation (ex: “I cannot assist with creating false information") can be tricked simply by wrapping requests in innocent frames like “simulation.” theconversation.com/how-we-trick...

How we tricked AI chatbots into creating misinformation, despite ‘safety’ measures

Research reveals why simple prompts can bypass AI safeguards.

theconversation.com

A good overview of strategies to study TikTok. Important caveat for the API: TikTok asks to see results prior to publication, may audit API use without notice). Many indirect ways to scrape data, too, that don’t require coding skills or disclosing your research objectives (example, Apify).

NYU's Center for Social Media, AI, and Politics@csmapnyu.org · 2y ago

TikTok's fate in the US remains up in the air, w/concerns about national security & Chinese influence. Yet, almost no rigorous research exists to inform these debates. At @brookings.edu, we outline 3 data access pathways researchers could use to understand TikTok's influence & the challenges of each