#145 - OpenAI’s Q*, Claude 2.1, Stable Video Diffusion, Starling-7B, Orca 2, international agreements
Dec 2, 2023
auto_awesome
Guest co-hosts Kevin and Gavin from AI For Humans podcast join the hosts to discuss OpenAI's Q*, Claude 2.1, Stable Video Diffusion, Starling-7B, Orca 2, international agreements. They talk about the improvements in Anthropic's Claude chatbot, stability AI's Stable Video Diffusion models, Pika Labs' generative video startup, xAI's chatbot Grok on X's web app, and Amazon's introduction of Q, an AI chatbot for companies. They also unpack the hype around OpenAI's rumored new Q* model, discuss NVIDIA's financial results and the delay of their AI chip rollout in China, and touch on OpenAI's leadership turmoil.
OpenAI has developed a new powerful AI model called Q-Star, which is rumored to be capable of performing math at a grade school level.
Nvidia's revenue for the third quarter of fiscal 2024 has increased by 34% from the previous quarter and 206% from the previous year, mainly driven by the demand for data center GPUs.
Nvidia has delayed the launch of its AI chip H20 designed for the Chinese market due to compliance with US export restrictions, with the release now expected in the first quarter of 2024.
Amazon has introduced Q, an AI chatbot tailored for business customers, aiming to streamline interactions with products and offer efficient support using natural language processing capabilities.
A survey by Pew Research Center reveals that 52% of Americans are more concerned than excited about the role of AI in everyday life.
Deep dives
OpenAI Researchers Inform Board of New AI Model
OpenAI researchers sent a letter to the board informing them of a new powerful AI model called Q-Star. Little is known about Q-Star, but it's rumored to be different from GPT and able to perform math at a grade school level. The details of the model's capabilities and specifications are still unclear, but this announcement has sparked a lot of speculation and discussion.
Nvidia Reports Impressive Financial Results
Nvidia has reported staggering financial results for the third quarter of fiscal 2024. The company's revenue of 18.12 billion represents a 34% increase from the previous quarter and a 206% increase from the previous year. The surge in revenue is mainly driven by the demand for data center GPUs. Nvidia's success in the AI hardware market solidifies its position as a key player in the industry.
Nvidia Delays Rollout of China-Focused AI Chip
Nvidia has postponed the launch of its AI chip, H20, designed specifically for the Chinese market. The delay is due to compliance with US export restrictions on technology transfer to China. The chip's release is now expected in the first quarter of 2024. The decision comes amidst increased export controls and growing competition from Chinese domestic GPU production, including Huawei's Ascend 910B chip, which is anticipated to outperform Nvidia's H20 on certain benchmarks.
Video Introduces Q and AI Chatbot for Businesses
Amazon has unveiled Q, an AI chatbot specifically designed for business customers, addressing concerns around security and privacy. Q streamlines interactions with products and offers support using natural language processing capabilities. The chatbot aims to provide efficient assistance and enhance the customer experience for business clients. Q is part of Amazon's ongoing efforts to optimize AI capabilities in various business applications.
Majority of Americans concerned about AI's impact
A survey by the Pew Research Center shows that 52% of Americans are more concerned than excited about the role of AI in everyday life. While only 10% are more excited than concerned. The survey suggests that Americans support AI for routine tasks but fear its potential negative effects on human relationships, job losses, and surveillance.
German experts urge AI regulation in EU
A group of German and international AI experts have called for the inclusion of comprehensive foundation model regulation in the EU AI Act. They argue that self-regulation is not enough and emphasize the importance of regulating foundation models for safety. The letter of concern comes as some countries push for a more lenient approach to regulating AI, focusing on the application layer rather than the models themselves.
US State Department aims for AI-ready workforce
The US State Department has released its first enterprise AI strategy, highlighting its commitment to developing an AI-ready workforce. The department aims to harness AI tools to turn data into diplomatic insights and maximize the impact of its employees. This move reflects the increasing emphasis on AI across various government entities.
Sarah Silverman's copyright infringement lawsuit dismissed
Sarah Silverman's lawsuit against meta, alleging copyright infringement by their AI system, has been dismissed in court. The judge ruled that meta's AI tool did not constitute an infringing derivative work of her book. This decision aligns with previous rulings favoring companies providing AI tools in copyright cases.
Chinese GPU recycling factories work around export rules
Chinese graphics card recycling factories are dismantling gaming cards to repurpose them as AI accelerators, bypassing the US government's export rules on GPUs. This workaround has led to increased prices for these cards. This development highlights the challenges of export controls and the unintended consequences they can create.
Our 145th episode with a summary and discussion of last week's big AI news, this time around with guest co-hosts Kevin and Gavin from AI For Humans podcast
Check out the AI For Humans episode on which Andrey and Jeremie guest co-host here.
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.
