AI Unveiled (Audio)
AI Unveiled (Audio)
English | 2023 | 9780138364885 | 3gp@64 kbps | 1h 28m | 41 MB
AI Unveiled
Hosted by Sinan Ozdemir
Listen to expert Sinan Ozdemir demystify AI and its technical, ethical, and legal impacts
Series Description
AI Unveiled, hosted by Sinan Ozdemir, is an in-depth audio series that explores the multifaceted realm of artificial intelligence. The show guides listeners through the technical intricacies of AI, the ethical dilemmas, and the legal implications of AI training data, touching upon issues like copyright infringement and intellectual property rights. Not shying away from the controversies, the series also investigates the line between moderation and censorship in AI outputs. Join Sinan as he pulls back the curtain on the challenges and ethical quandaries inherent in developing, fine-tuning, and deploying AI, all while shedding light on the future opportunities and risks.
Related Learning
Book: Quick Start Guide to Large Language Models
Video: Quick Guide to ChatGPT, Embeddings and Other Large Language Models
Video: Introduction to Transformer Models for NLP
Live Training:https://learning.oreilly.com/search/?q=sinan%20ozdemir&type=live-event-series
Episode 1 Complete the sentence—Introduction to LLMs and Chat___
Host Sinan Ozdemir talks through the evolution and impact of language models like ChatGPT, exploring their technological underpinnings, ethical challenges, and transformative role in search engines, coding, and more.
Episode 2 Critical Reasoning—A look at the current state of LLMs
Sinan Ozdemir explores the ethical and practical limitations of Large Language Models like ChatGPT. He questions who gets to build and control these powerful tools, while also shedding light on their inability to perform tasks requiring deep critical reasoning or moral judgment.
Episode 3 The haves and the have nots—Training LLMs
Sinan Ozdemir delves into the financial disparities in the world of Large Language Models. Big tech firms can spend millions to develop advanced models, putting mid-sized and small companies at a disadvantage. However, the rise of open-source initiatives and educational resources offers some hope for leveling the playing field. The episode emphasizes the need for more equitable access to AI technologies.
Episode 4 Perfect, The enemy of good—Evaluating LLMs
Sinan Ozdemir discusses the challenges of evaluating Large Language Models. He highlights the limitations and advantages of two key methods: offline evaluation, which uses past data, and online evaluation, which assesses real-world impact. The episode calls for more comprehensive, hybrid evaluation methods to ensure models are effective and ethical.
Episode 5 As an AI language model I've been aligned with human feedback
Sinan Ozdemir explores the critical topic of AI alignment, focusing on methods like RLHF and RLAIF. He discusses OpenAI's InstructGPT as a milestone in making large language models more attuned to human instructions. Sinan also addresses the ethical and financial implications of AI alignment, pointing out its importance not just for industry giants but for small-scale developers as well.
Episode 6 Who said that?—The hidden biases in LLMs
Sinan Ozdemir delves into the urgent issue of biases in Large Language Models, using real-world examples from text-to-image generation models like Stable Diffusion and ERNIE-ViLG. He explores how cultural values manifest in these models and questions the ethical implications of Western-centric training data. The episode also tackles the emerging legal challenges facing AI giants like OpenAI and Meta, including issues of copyright infringement and enterprise liability. The conversation underscores the complexity of biases in AI and prompts listeners to consider the ethical and legal ramifications.
Episode 7: Rise of open source and data is finally the new oil
Sinan Ozdemir dives into Open-Source LLMs and the massive datasets that power these technological powerhouses. Open source Large Language Models (LLMs) play a pivotal role in the democratization of AI and data. With dozens of new free and fully available models released to the public, these free and open source models let anyone train their own AI systems without having to rely on another company. By sharing model weights and training methods, these open source projects enable a broader community to contribute to and benefit from AI advances, accelerating progress and ensuring that the power of AI does not remain siloed in the hands of a few. He also discusses what fuels these powerful engines of innovation, the massive datasets that they learn from.
Episode 8: SAG AFTRA, Netflix's prompt engineering job, Biden's Executive order on AI
Sinan Ozdemir discusses the latest Executive Order on AI and the historic SAG-AFTRA strike. The comprehensive order calls for the creation of guidelines, standards, and best practices, all designed to steer the ship of AI innovation in a direction that safeguards our collective future. SAG-AFTRA's historic strike placed a glaring spotlight on the deep-seated concerns surrounding AI in the entertainment domain. These events underscore the importance of ethical regulations surrounding AI with an emphasis on individual rights and the economic implications of AI.
Episode 9: Now you see me—Multimodality
Sinan Ozdemir explains Multimodal AI, key models in the space, and how these models work. Multimodal AI refers to AI systems that can understand and process different types of data inputs—such as text, sound, images, and videos—to perform tasks that traditionally required human intelligence. Sinan explains how models like DALL-E, Stable Diffusion, and ERNIE-ViLG paint images from text and how models like Wav2lip can merge together audio and video.. He also comments on the ethical considerations, rights, and the very human concerns surrounding use of these models.
Sinan Ozdemir is founder and CTO of LoopGenius, where he uses state-of-the-art AI to help people create and run their businesses. He has lectured in data science at Johns Hopkins University and authored multiple books, videos and numerous online courses on data science, machine learning, and generative AI. He also founded the recently acquired Kylie.ai, an enterprise-grade conversational AI platform with RPA capabilities. Sinan most recently published Quick Guide to Large Language Models, and launched a podcast audio series, AI Unveiled. Ozdemir holds a master's degree in pure mathematics from Johns Hopkins University.
DOWNLOAD
https://rapidgator.net/file/e8ede7fb7f75963b6c0a9763874d9e5a/AI_Unveiled_(Audio).rar.html
https://nitroflare.com/view/1F411E6530D90E2/AI_Unveiled_%2528Audio%2529.rar
https://nitroflare.com/view/1F411E6530D90E2/AI_Unveiled_%2528Audio%2529.rar
