Podcast Summary
Hollywood's longest strike: SAG vs studios over actors' digital likenesses: The use of actors' digital scans in AI is a contentious issue, with labor unions demanding compensation and consent and studios proposing indefinite use without. This dispute underscores the need for clear guidelines as AI becomes more prevalent.
During the longest Hollywood strike in history, the Screen Actors Guild (SAG) and studios clashed over the use of actors' digital likenesses in AI. The studios proposed scanning actors once and using their digital scans indefinitely without consent or compensation. SAG, on the other hand, demanded compensation and consent for the use of their members' scans. This dispute highlights the potential power struggles and complexities surrounding AI use and the rights of individuals, particularly in the context of organized labor. The agreement's specifics have not been disclosed yet, but this situation underscores the importance of addressing these issues as AI becomes increasingly prevalent.
Geopolitical tensions and export restrictions fuel the development of AI chips tailored to specific markets: Geopolitical tensions result in export restrictions on AI chips, driving companies to create market-specific chips and investors to embrace AI for managing large portfolios, while safety concerns lead to the launch of new labs to address potential issues.
The race for advanced AI technology has led to geopolitical tensions, specifically between the US and China, resulting in export restrictions on AI chips. These restrictions have forced companies like Nvidia to develop new chips tailored to the Chinese market while adhering to US regulations. Meanwhile, the investment world is also embracing AI, with Norway's Sovereign Wealth Fund, the world's largest, using AI to manage its investments and boost productivity. This trend is expected to spread among other investors, making AI an essential tool for managing vast portfolios. Additionally, safety concerns surrounding AI have led to the launch of the Safety Evaluations and Analysis Lab (SEAL) by Scale AI to address these issues. These developments underscore the growing importance of AI in various sectors and the ongoing efforts to navigate its implications.
Companies focus on responsible AI development and compliance: Scale AI launches lab to evaluate AI risks and biases, driven by ethical concerns and regulatory needs. Amazon reportedly working on new model Olympus, while Humane's AI pin gains buzz. AI-crypto intersection grows, check out Web 3 with 16z Crypto podcast.
Companies are increasingly focusing on responsible AI development and deployment, driven in part by upcoming regulations and compliance needs. Scale AI is launching a new lab to help their customers evaluate potential risks and biases in AI software, and this effort seems to be motivated by both a desire for ethical AI and the need to stay ahead of regulatory requirements. Meanwhile, there's buzz around Humane's AI pin, a new device reportedly costing $699 with a $24 monthly subscription. In the world of AI, Amazon is reportedly working on a new model called Olympus, which is said to have twice the parameters of GPT 4. The intersection of AI and crypto is a growing trend, and for those interested in this topic, Web 3 with a 16z Crypto is a recommended podcast.
Amazon developing new AI model Olympus with more parameters than GPT 4: Amazon is creating a new AI model, Olympus, with more parameters than GPT 4 to enhance their AWS offerings and compete with tech giants in foundation and frontier models, while also providing developers with choice and customization through their service.
Amazon is developing a new AI model named Olympus, which is rumored to have twice as many parameters as OpenAI's GPT 4. This move is part of the larger competition between tech giants in the field of foundation and frontier models. Amazon, already offering AI services through their Bedrock platform, believes having their unique model within the AWS ecosystem could be an attractive offering for their customers. However, Amazon's previous attempt to debut a similar LLM named Bedrock at their annual customer conference had to be postponed due to technical issues. When ChatGPT was released and went viral, Amazon recognized the need to adapt and instead of leading with a model-first strategy, they built a service that allows developers to connect with various large language models, emphasizing choice and customization for enterprise customers.
Amazon's New AI Model, Olympus, Faces Skepticism: Amazon invests in ethical AI development and enters AI race with Olympus, but success depends on performance and capabilities
Amazon's upcoming AI model, Olympus, faces skepticism due to mixed reviews of its predecessor Titan and the perceived advancements of competitors like OpenAI. While Amazon has the resources to compete in terms of model size, it's not guaranteed that a larger model is better, as Meta's chief AI scientist Jan Lacun pointed out. The effectiveness of a model's parameters depends on how they're utilized. Amazon's investment in anthropic, a company focused on aligning AI with human values, could be seen as a hedge but also a strategic move to ensure ethical development of AI technology. The pressure to launch a state-of-the-art model for powering Alexa and customizing AI for every book, author, and topic in its bookstore further emphasizes the importance of Amazon entering the AI race. The potential debut of Olympus at re:Invent could help alleviate concerns about its competitiveness, but the model's success ultimately depends on its performance and capabilities.
Google's delayed release of advanced AI model Gemini raises industry concerns: Google's anticipated advanced AI model Gemini, seen as a potential game-changer, faces delays, raising industry concerns about the company's progress and competitiveness in the generative AI market
The race to release advanced generative AI models is heating up, with companies like Microsoft and Anthropic forming partnerships to access a wider range of models, while Google's anticipated Gemini model remains delayed. The significance of Gemini goes beyond the industry, as it's seen as a potential game-changer that could surpass the capabilities of current models like GPT-4. However, Google's absence from the generative AI scene has raised questions and concerns, with some commentators expressing surprise and disappointment at the company's perceived lack of progress. The longer Gemini remains in development, the higher the stakes become for Google to deliver a model that meets or exceeds expectations, particularly in terms of cost, speed, and functionality. The delay in Gemini's release, coupled with Google's regulatory and organizational challenges, has fueled speculation and criticism from the industry.
Google's Gemini model vs OpenAI's Gobi: The race for advanced language models: Google's Gemini and OpenAI's Gobi are the latest advanced language models expected to be released, with Gemini predicted to surpass GPT 4's benchmark score and focus on solving hallucinations using the search graph and mixture of experts output, while Gobi is predicted to be announced in October 2024 and aim for even higher performance.
The race for advanced language models is heating up, with several major players, including Google, Meta, Amazon, OpenAI, and Elon Musk's XAI, vying for dominance. According to prediction markets, Google's Gemini model is expected to be released soon, possibly this quarter or in Q1 2024. Metaculus predicts a benchmark score of 90.13 for Gemini on release, which would surpass GPT 4's score of 86.5. Google is aiming for Gemini to position itself between existing models by focusing on solving hallucinations using the search graph and mixture of experts output. Meanwhile, OpenAI is also working on a more advanced model, Gobi, which is predicted to be announced in about 11 months, on October 5, 2024. The speed of innovation in the field is increasing rapidly, with models like Grok being developed and released in just a few months. Amazon's Olympus has a significant challenge ahead in entering this competitive landscape, but it is expected to officially announce its model at some point. The race for advanced language models is intensifying, and the only question is which company will come out on top and when. Stay tuned for updates on this developing story.