Last Week in AI #157 - Gemini controversy, new Mistral models, Deepmind's Genie & Griffinn, AI Warfare is here
61 snips
Mar 3, 2024 The podcast dives into the heated controversy over Google's Gemini, which mistakenly generated racially diverse Nazis. Innovations in AI are spotlighted, like Stability's Stable Diffusion 3 and Mistral's chat assistant, as competition heats up in the market. Exciting multimedia advancements feature Windows’ new Magic Eraser and Adobe's generative audio tools. Discussions also cover Microsoft's strategic investment in Mistral and the rising pressures of AI warfare, highlighting both the potential and ethical dilemmas that come with rapid technological evolution.
AI Snips
Chapters
Books
Transcript
Episode notes
Gemini's Nazi Controversy
- Google's Gemini image generator produced racially diverse Nazis, even when prompted with historically accurate contexts.
- This led to widespread criticism, a stock drop, and the feature's temporary removal.
Technical and Cultural Issues at Google
- Gemini's blunder highlights the limitations of current AI alignment techniques and potential cultural issues within Google.
- The product's release despite obvious flaws suggests a problematic internal culture.
DALL-E-2's Similar Diversity Issue
- DALL-E-2 had a similar issue with forced diversity, modifying user prompts to include specific racial or gender characteristics.
- This suggests a common, ham-fisted approach to addressing bias in image generation models.
