Episode 39

On LLMs, feat Eugene Cheah | TRACES Appendix 39

In this engaging conversation, Eugene and Cristian delve into the world of open-source AI models. They explore the intricate challenges of training and operating large language models, highlighting the delicate balance between regulation and innovation in the AI industry. The discussion offers insights into the complexities of AI development, emphasizing the need for thoughtful dialogue and collaboration to navigate the potential risks and benefits. The conversation also addresses the potential dangers of deepfake technology and underscores the importance of good faith efforts in the regulation of AI. Eugene and Cristian discuss various critical topics, such as the necessity of supporting multiple languages in AI models, the limitations of current benchmarks, and the impressive performance enhancements of RWKV compared to traditional Transformers. Furthermore, the role of humor in language models is examined, shedding light on how AI can be both a technical marvel and a tool for creativity. Join Eugene and Cristian as they provide a comprehensive overview of the current landscape and future directions of AI technology, offering valuable perspectives for anyone interested in the field. Find me at cristian@ccb.life PRE-ORDER TRACES: A PSY-FI NOVEL NOW (https://ccblife.gumroad.com/l/traces) Also, who are you? Get a draft of TRACES if you fill out this form (https://forms.gle/rFnVFrCNUAJz7Fvn7) About the Guest: Eugene is a repeat founder with multiple successful exits. Before his entrepreneurial ventures, he developed enterprise software solutions for all major banks and insurance companies throughout Singapore. For the past decade, Eugene has been actively contributing to the open-source community across various domains. His work spans from infrastructure projects like Kubernetes and Rancher to GPU coding with gpu.js, and more recently, advancements in the AI field with rwkv. Currently, Eugene is focused on making AI accessible to everyone, in every language. Set-Up: - Camera: https://amzn.to/3PZVscb (don't laugh) - Microphone: https://amzn.to/46f3pB5 - Teleprompter Stand: https://amzn.to/3tgS98y - Telepromter App: https://amzn.to/46jdH31 - Teleprompter Screen: https://amzn.to/3PNfKFI (yup) - Headphones: https://amzn.to/46gMSwo Timestamps 00:00 Introduction and Background 03:16 The AI Race and Different Approaches 06:50 Pruning Giant Models and Delegating Solutions 20:54 Forming Alliances and Building Talent Pools 27:59 The Interplay Between Big Tech and Regulation 34:53 Balancing Regulation and Innovation in AI 44:00 Regulation of Language Models 45:02 Including Multiple Languages 51:44 Limitations of Benchmarks 01:20:23 The Difficulty of Mastering Humor Hashtags #AI #ArtificialIntelligence #OpenSourceAI #LanguageModels #Deepfake #AIRegulation #AIInnovation #RWKV #TransformersAI #MultipleLanguages #AIDevelopment #AIChallenges #AITraining #AIConversation #TechDialogue #AIInsights #AIrisks #DeepfakeTechnology #AIRegulationDialogue #BenchmarksAI #PerformanceImprovements #AIHumor #AIandCreativity #AIIndustry #AIFuture