Suspicion machines ⚙️
from Practical AI Hosted by Daniel Whitenack and Chris Benson
In this enlightening episode, we delve deeper than the usual buzz surrounding AI’s perils, focusing instead on the tangible problems emerging from the use of machine learning algorithms across Europe. We explore “suspicion machines” — systems that assign scores to welfare program participants, estimating their likelihood of committing fraud. Join us as Justin and Gabriel share insights from their thorough investigation, which involved gaining access to one of these models and meticulously analyzing its behavior. Sponsors: Changelog News – A podcast+newsletter combo that’s brief, entertaining & always on-point. Subscribe today. Fastly – Our bandwidth partner. Fastly powers fast, secure, and scalable digital experiences. Move beyond your content delivery network to their powerful edge cloud platform. Learn more at fastly.com Fly.io – The home of Changelog.com — Deploy your apps and databases close to your users. In minutes you can run your Ruby, Go, Node, Deno, Python, or Elixir app (and databases!) all over the world. No ops required. Learn more at fly.io/changelog and check out the speedrun in their docs. Featuring: Gabriel Geiger – X Justin-Casimir Braun – X Daniel Whitenack – Website, GitHub, X Show Notes: Article - “Inside the suspicion machine” The methodology behind Justin and Gabriel’s report Upcoming Events: Register for upcoming webinars here!
Embed this episode
Chapters
- 0:07 Welcome to Practical AI
- 0:43 Suspicion machines
- 4:14 Signs of bias
- 7:11 How to start investigating
- 9:29 Predictive analytic trends in gov.
- 11:37 Can you justify using these systems?
- 14:24 Requesting Rotterdam's data
- 16:47 The model's capabilities
- 20:23 Important features
- 23:15 Understanding the labels
- 26:50 Sponsor: Changelog News
- 28:50 Contract timeline
- 31:19 How effective is the model?
- 34:21 Public perception and reactions
- 39:57 Hope for the future
- 45:29 Goodbyes
- 46:09 Outro
No transcript yet.