Chris Albon — ML Models and Infrastructure at Wikimedia
Gradient Dissent: Conversations on AI
00:00
Detecting the Most Common Type of Spam Attacks on Machine Learning Systems
The most common one is definitely like adding swear words, it's something, you know, so as you as you can imagine. You know the ones that are, are the most dangerous are definitely the ones that the attackers have a lot of resources. Everything we do is open source, you can see my internal chat,You can see my ticket existence like my Jira is totally public. Like what I'm working on in a given day is public I'm live streaming the work that I'm doing every single like other week or something like that. All this is open, and every single article on adversarial attacks has been written about how to deal with them.
Transcript
Play full episode
