#157 - Gemini controversy, new Mistral models, Deepmind's Genie & Griffinn, AI Warfare is here

Published: March 3, 2024, 8:05 p.m.

Our 157th episode with a summary and discussion of last week's big AI news!\nCheck out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there\u2019s a video version on YouTube.\nBonus plug: also check out this new book by Stanford AI expert, bestselling author, and Last Week in AI supporter Jerry Kaplan! Generative Artificial Intelligence: What Everyone Needs to Know \nRead out our text newsletter and comment on the podcast at https://lastweekin.ai/\nEmail us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai\nTimestamps + links:\n(00:00:00) Intro / Banter\nTools & Apps\n(00:04:55) Google apologizes for \u2018missing the mark\u2019 after Gemini generated racially diverse Nazis\n(00:15:07) Stability announces Stable Diffusion 3, a next-gen AI image generator\n(00:19:04) Mistral AI releases new model to rival GPT-4 and its own chat assistant\n(00:24:56) Windows just got its own Magic Eraser to AI-modify your photos\n(00:25:47) Adobe previews new cutting-edge generative AI tools for crafting and editing custom audio\n(00:27:47) AI video wars heat up as Pika adds Lip Sync powered by ElevenLabs\n\nApplications & Business\n(00:30:09) Microsoft Strikes Deal with France\u2019s Mistral AI\n(00:33:45) Figure Raises $675M at $2.6B Valuation and Signs Collaboration Agreement with OpenAI\n(00:37:05) Nvidia posts record revenue up 265% on booming AI business\n(00:39:54) MediaTek\u2019s latest chipsets are now \u2018optimized\u2019 for Gemini Nano\n(00:41:28) Tumblr\u2019s owner is striking deals with OpenAI and Midjourney for training data, says report\n(00:43:45) Mistral AI models coming soon to Amazon Bedrock\n\nProjects & Open Source\n(00:44:34) Generative AI Startup Mistral Releases Free \u2018Open-Source\u2019 7.3B Parameter LLM\n(00:46:57) Google Delves Deeper Into Open Source with Launch of Gemma AI Model\n(00:51:10) Microsoft releases its internal generative AI red teaming tool to the public\n(00:54:25) Introducing Phind-70B \u2013 closing the code quality gap with GPT-4 Turbo while running 4x faster\n\nResearch & Advancements\n(00:57:51) Genie: Generative Interactive Environments\n(01:07:08) Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language Models\n(01:15:16) Quantum Circuit Optimization with AlphaTensor\n(01:20:56) Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs\n(01:22:10) Repetition Improves Language Model Embeddings\n\nPolicy & Safety\n(01:25:45) AI Warfare Is Already Here\n(01:29:36) Man admits to paying magician $150 to create anti-Biden robocall\n(01:32:45) Google DeepMind forms a new org focused on AI safety\n(01:34:53)\xa0 Facebook whistleblower, AI godfather join hundreds calling for deepfake regulation\nUS regulators investigate whether OpenAI investors were misled, say reports\n(01:36:51)\xa0 Users Say Microsoft's AI Has Alternate Personality as Godlike AGI That Demands to Be Worshipped\n\nSynthetic Media & Art\n(01:40:23) The Intercept, Raw Story, and AlterNet sue OpenAI and Microsoft\n(01:41:15) A viral photo of a guy smoking in McDonald's is completely fake \u2014 and of course made by AI\n\nFun!\n(01:42:54) Impossible AI Food