Last Week in AI
#174 - Odyssey Text-to-Video, Groq LLM Engine, OpenAI Security Issues
Podcast summary created with Snipd AI
Quick takeaways
- Innovative memory training enables real-time model adaptation, improving learning efficiency and scalability.
- AMD's strategic acquisition of Silo AI enhances software proficiency and innovation in the semiconductor industry.
- Tension between Sequoia and Andreessen Horowitz exposes differing AI chip investment strategies and competitive dynamics.
- Elon Musk plans to build a powerful AI training cluster with 100,000 H100 GPUs for XAI advancements.
- Intel faces challenges in constructing Fab 29 in Germany due to environmental protests and subsidy hurdles.
- Vimeo introduces AI content labels, Etsy updates seller policy, and Bumble enables reporting of AI-generated profiles for transparency and authenticity.
Deep dives
Innovative Memory Model for RNNs to Enhance Learning at Test Time
Introducing a new concept where hidden states in RNNs are treated as trainable models to adapt at test time, enabling continual learning and improved performance. This approach allows memory to be updated with self-supervised training at inference, leading to enhanced adaptability and scaling.
Pioneering AI Robotics Startup Skilled AI Secures $300 Million in Series A
Skilled AI, a Pittsburgh-based startup, attains a $1.5 billion valuation in its Series A funding round. Focused on developing cost-effective, general-purpose robots across various industries, the company aims to leverage top-tier investors and expertise to drive advancements in robotics technology.
Intel Initiates Construction on Cutting-Edge Chip Fab in Germany
Intel embarks on establishing Fab 29 in Germany, slated to be a leading chip fab in Europe, despite facing environmental objections and EU subsidy approvals. The facility aims to bolster chip fabrication capabilities critical for AI advancements and national security, positioning Intel in the competitive chip manufacturing landscape against industry giants like TSMC.
Intel to Acquire Silo AI in $665 Million Deal to Strengthen Software Expertise
AMD's acquisition of Finnish AI company Silo AI for $665 million underscores their strategic move to enhance software proficiency. With 100 PhDs and 300 employees aligned with Silo AI joining AMD, the acquisition aims to fortify AMD's software capabilities and offerings in a rapidly evolving semiconductor landscape.
Elon Musk Plans to Establish 100,000 NVIDIA GPU AI Cluster
Elon Musk reveals intentions to build a cutting-edge AI training cluster comprising 100,000 H100 GPUs, marking a significant step to boost XAI's computing power. Despite parting ways with Oracle, XAI's mission to leverage vast GPU resources to enhance training capabilities remains steadfast.
Groundbreaking AI Model Incorporates Dynamic Hidden State Training at Inference Time
Significant AI model innovation involves adapting hidden states through trainable models during inference time, revolutionizing learning capabilities by continuously updating memory with new data. The dynamic memory adaptation approach yields impressive performance enhancements and fosters a more adaptable AI framework.
AI Chip Supplier Clashes Continue Between Sequoia and Andreessen Horowitz
Tension arises between Sequoia and Andreessen Horowitz over AI chip stockpiling strategies amidst the Gen AI boom, highlighting differing investment perspectives. The clash reflects contrasting viewpoints on scaling and profitability in the AI chip sector, shedding light on competitive dynamics and investment strategies among top VCs.
German Chip Fab Construction Stalled Amid Environmental Concerns and EU Subsidy Delays
Intel faces roadblocks in Fab 29 construction in Germany due to environmental protests and EU subsidy hurdles, delaying progress to establish a leading chip fab in Europe. The facility's strategic significance for chip fabrication capabilities and national security underscore challenges around environmental considerations and subsidy approvals.
Unique Memory Training Approach Enhances Learning and Scalability in AI Models
Innovative memory training technique revolutionizes learning by enabling inference-time model training through continuous memory updating. This dynamic approach optimizes learning efficiency and scalability, showcasing enhanced model adaptability and performance across diverse AI tasks.
AMD Acquisition of Silo AI Strengthens Software Expertise in $665 Million Deal
AMD's strategic acquisition of Finnish AI company Silo AI for $665 million elevates their software capabilities to drive innovation in the semiconductor landscape. The integration of top talent from Silo AI bolsters AMD's software proficiency and advances their offerings in the competitive semiconductor market.
Vimeo Launches AI Content Labels to Prevent Confusion
Vimeo has introduced new AI content labels to help creators disclose AI-generated content, ensuring transparency and reducing confusion among viewers.
Startup Avail Aims to Assist Media Licensing for AI Training
Avail, a tech startup, is launching Corpus, a product to help media companies and independent creators license their content for AI model development, providing an opportunity for creators to monetize their data.
Etsy Introduces AI Content Guidelines in New Seller Policy
Etsy has updated its seller policy to include guidelines for labeling AI-generated content, requiring sellers to identify AI involvement in creating items, ensuring transparency and limiting deceptive practices on the platform.
Bumble Users Can Report Profiles with AI-Generated Photos
Bumble has enabled users to report profiles containing AI-generated photos, aiming to combat the use of fake profiles, scammers, and deceptive practices on the dating app, prioritizing user protection and authenticity.
Our 174rd episode with a summary and discussion of last week's big AI news!
With hosts Andrey Kurenkov (https://twitter.com/andrey_kurenkov) and Jeremie Harris (https://twitter.com/jeremiecharris)
In this episode of Last Week in AI, we delve into the latest advancements and challenges in the AI industry, highlighting new features from Figma and Quora, regulatory pressures on OpenAI, and significant investments in AI infrastructure. Key topics include AMD's acquisition of Silo AI, Elon Musk's GPU cluster plans for XAI, unique AI model training methods, and the nuances of AI copying and memory constraints. We discuss developments in AI's visual perception, real-time knowledge updates, and the need for transparency and regulation in AI content labeling and licensing.
See full episode notes here.
Read out our text newsletter and comment on the podcast at https://lastweekin.ai/
If you would like to become a sponsor for the newsletter, podcast, or both, please fill out this form.
Email us your questions and feedback at contact@lastweekinai.com and/or hello@gladstone.ai
Timestamps + links:
- (00:00:00) Intro AI Song
- (00:00:41) Pre News Banter
- Tools & Apps
- (00:07:09) Odyssey Building 'Hollywood-Grade' AI Text-to-Video Model to Compete With Sora, Gen-3 Alpha
- (00:10:28) Anthropic’s Claude adds a prompt playground to quickly improve your AI apps
- (00:15:06) Figma pauses its new AI feature after Apple controversy
- (00:18:30) Quora’s Poe now lets users create and share web apps
- (00:20:54) Suno launches iPhone app — now you can make AI music on the go
- Applications & Business
- (00:21:42) Groq unveils lightning-fast LLM engine; developer base rockets past 280K in 4 months
- (00:27:03) Microsoft and Apple ditch OpenAI board seats amid regulatory scrutiny
- (00:29:39) OpenAI and Arianna Huffington are working together on an ‘AI health coach’
- (00:33:38) AI coding startup Magic seeks $1.5-billion valuation in new funding round, sources say
- (00:37:01) Sequoia and Andreessen Horowitz Clash Over AI Chip Supplies Amid Gen AI Boom
- (00:43:30) Elon Musk Reveals Plans To Make World’s “Most Powerful” 100,000 NVIDIA GPU AI Cluster
- (00:46:25) AMD plans to acquire Silo AI in $665 million deal
- (00:48:00) AI robotics startup raises US$300 million, including from Jeff Bezos
- (00:52:11) Intel begins groundwork on Magdeburg chip fab despite 13 remaining regulatory and environmental objections
- Research & Advancements
- (00:55:21) Learning to (Learn at Test Time): RNNs with Expressive Hidden States
- (01:03:12) Data curation via joint example selection further accelerates multimodal learning
- (01:09:11) CopyBench: Measuring Literal and Non-Literal Reproduction of Copyright-Protected Text in Language Model Generation
- (01:13:25) Just read twice: closing the recall gap for recurrent language models
- (01:15:25) CodeUpdateArena: Benchmarking Knowledge Editing on API Updates
- (01:18:31) Composable Interventions for Language Models
- (01:24:09) Mind-reading AI recreates what you're looking at with amazing accuracy
- Policy & Safety
- (01:26:49) Covert Malicious Finetuning
- (01:31:23) OpenAI’s week of security issues
- (01:36:39) Here’s how OpenAI will determine how powerful its AI systems are
- (01:39:56) Me, Myself and AI: The Situational Awareness Dataset for LLMs
- (01:44:34) Exclusive: OpenAI partners with Los Alamos to study AI in the lab
- (01:47:36) Judge dismisses coders’ DMCA claims against Microsoft, OpenAI and GitHub
- (01:49:55) A former OpenAI safety employee said he quit because the company's leaders were 'building the Titanic' and wanted 'newer, shinier' things to sell
- Synthetic Media & Art
- (02:02:05) Outro + AI Song
Remember Everything You Learn from Podcasts
