Get email alerts on the Mitigation Tools topic
No spam. Unsubscribe anytime.
FLCC presenter recommends practical bias mitigations: Nightshade, Latimer, NotebookLM and prompting
Summary
Gadoo recommended a toolkit approach: defensive image tools (Nightshade), inclusive models (Latimer), source‑controllable assistants (NotebookLM), running LLMs locally where feasible, and human review plus careful prompting to reduce bias risks.
In the webinar, Gadoo outlined practical steps for creators and organizations to reduce bias. For image creators he described Nightshade, a tool that perturbs image pixels to prevent inclusion in large training corpora; for model selection he highlighted 'Latimer' as an example of an inclusive‑focused commercial model.
He also recommended NotebookLM for source‑controllable research workflows, advised running local LLMs where privacy and control are priorities, and reiterated simple user practices: do not put PII in prompts, use targeted prompting to avoid stereotyped outputs, and review AI outputs before sharing. Ortloff and Gadoo encouraged attendees to sign up for follow‑up sessions and resources.
AI generated
The text on this page is AI generated. Summaries, highlights, analysis, and video transcripts are all produced from the original source material.
AI can make mistakes, so if you spot one, and we will fix it for everyone.
Note: the source content is unaltered by us. Any content source we link to, be it a video, an audio recording, or a document, is presented exactly as its publisher released it. That publisher is usually a government body, sometimes an individual official or another organisation.
