Google's Gemini, a multimodal chatbot, is discussed including its launch, capabilities, and controversy surrounding its demo video. The podcast also covers Google's AlphaCode 2, Meta's Audiobox, Elon Musk's Grok AI plagiarism, and Gmail's AI-powered spam detection. The competition between NVIDIA and AMD in the AI hardware market is highlighted, along with discussions on Tesla's Dojo, Assembly AI funding, and Leonardo dot AI's fundraising. The chapter also explores the architecture and parameters of large language models, techniques to improve information recall, preventing jail breaks in chat GPT, and the EU AI Act and its implications.
Read more
AI Summary
Highlights
AI Chapters
Episode notes
auto_awesome
Podcast summary created with Snipd AI
Quick takeaways
Google's Gemini is a new AI chatbot that is facing mixed reviews in terms of its performance and authenticity.
Meta (formerly Facebook) has unveiled Audiobox, an AI that can clone voices and generate ambient sounds.
Elon Musk's AI company, Grok AI, was caught plagiarizing OpenAI's ChatGPT, leading to concerns among his fans.
Deep dives
Assembly AI raises $50 million in funding
Assembly AI, an applied AI venture, has raised $50 million in funding, bringing their total capital raise to $115 million. The company's customer base has grown 200% from last year, and their AI platform is handling approximately 25 million API calls per day. Assembly AI specializes in speech-related activities, providing AI speech-to-text models.
Generative AI art platform Leonardo.AI raises $31 million
Leonardo.AI, a Sydney-based generative AI art platform, has raised $31 million in funding. The platform has gained 7 million users and has generated over 700 million images. They recently launched an enterprise version, expanding their offerings in the AI art production market.
Legal Case Proceeds in Getty vs. Stability AI Copyright Infringement
The high court in England and Wales has ruled that the legal case between Getty Images and Stability AI can proceed. Getty Images has sued Stability AI for copyright infringement, database right infringement, trademark infringement, and more in relation to the use of copyrighted images in training the Stable-D Fusion model.
G7 Agrees on Guidelines for Generative AI
The G7, a group of seven major countries, has agreed on international guidelines for generative AI. The guidelines cover issues like misinformation and are part of a voluntary code of conduct for AI developers.
Top Executives Fired at Sports Illustrated Publisher After AI Debacle
Two senior executives, CEO Andrew Kraft and President Rob Barrett, have been fired from the publisher of Sports Illustrated following an AI-generated content controversy. The controversy involved AI-generated articles with fake offers and poorly executed aspects, leading to pushback and the termination of the executives.
MIT Group Releases White Papers on AI Governance
A group of MIT leaders and scholars has released a set of policy briefs outlining a framework for AI governance, primarily in the US. The papers address topics such as auditing new AI tools, liability, and regulatory oversight, with a focus on ensuring safety and thriving AI development.
Our 147th episode with a summary and discussion of last week's big AI news!
Correction: Gemini also supports audio
Also check 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’s a video version on YouTube.