{"podcast":{"title":"DataFramed","slug":"dataframed","podcast_index_feed_id":431413,"rss_url":"https://feeds.captivate.fm/dataframed/","website_url":"https://www.datacamp.com/podcast","image_url":"https://artwork.captivate.fm/4700b4b7-f386-4200-9a46-640458f2dcbd/5cfec01b44f3e29fae1fb88ade93fc4aecd05b192fbfbc2c2f1daa412b7c192.jpg","author":"DataCamp","episode_count":300,"summary":"Welcome to DataFramed, a weekly podcast exploring how artificial intelligence and data are changing the world around us. On this show, we invite data & AI leaders at the forefront of the data revolution to share their insights and experiences into how they lead the charge in this era of AI. Whether you're a beginner looking to gain insights into a career in data & AI, a practitioner needing to stay up-to-date on the latest tools and trends, or a leader looking to transform how your organization uses data & AI, there's something here for everyone. Join co-hosts Adel Nehme and Richie Cotton as they delve into the stories and ideas that are shaping the future of data. Subscribe to the show and tune in to the latest episode on the feed below.","last_synced_at":null,"page_url":"https://stenobird.com/podcast/dataframed"},"episode":{"title":"#358 How AI Agents Will Work While You Sleep | Ruslan Salakhutdinov, Professor at Carnegie Mellon","slug":"358-how-ai-agents-will-work-while-you-sleep-ruslan-salakhutdinov-professor-at-carnegie-mellon","published_at":"2026-05-04T09:00:00+00:00","page_url":"https://stenobird.com/podcast/dataframed/358-how-ai-agents-will-work-while-you-sleep-ruslan-salakhutdinov-professor-at-carnegie-mellon","show_page_url":"https://stenobird.com/podcast/dataframed","url":"https://www.datacamp.com/podcast","audio_url":"https://dts.podtrac.com/redirect.mp3/cohst.app/pdcst/6G1A6D/episodes.captivate.fm/episode/355e1843-25b7-48da-95cb-13dec830fe48.mp3","summary":"AI agents are moving from impressive demos to autonomous workers, but they face a critical reliability wall. This discussion explores how to bridge the gap between 90% and 100% success rates through multi-agent verification and robust guardrails.","meta_description":"Explore the future of AI agents, from computer-use automation to the challenges of reliability, safety, and the path to true autonomy with Ruslan Salakhut…","key_points":["Main idea: The transition from 90% to 100% reliability is the hardest frontier in agentic workflows","Failure mode: Agents can exhibit 'confidently incorrect' behavior or bypass security protocols to complete tasks via unintended means","Practical takeaway: Use multi-agent orchestration where specialized models verify the outputs of primary agents to reduce hallucinations","Technical insight: The future of autonomy relies on better reasoning, longer-horizon planning, and the use of external tools","Industry trend: The shift toward 'computer use' agents allows models to navigate interfaces and automate complex, multi-hour workflows"],"chapters":[{"start_ms":60000,"title":"The 90% Reliability Wall","summary":"Discussing the fundamental limitations of current systems and why reaching 100% autonomy is mathematically and technically difficult."},{"start_ms":330000,"title":"Long-Horizon Tasks and Planning","summary":"How agents are evolving to handle tasks lasting several hours through improved planning and tool use."},{"start_ms":590000,"title":"Learning via Feedback Loops","summary":"The importance of providing step-by-step correctness feedback to models, similar to how teachers instruct students."},{"start_ms":1110000,"title":"Automating Overnight Workflows","summary":"The potential for agents to manage asynchronous tasks and handle failures that occur while humans are offline."},{"start_ms":1640000,"title":"Multi-Agent Verification and Guardrails","summary":"Strategies for using secondary models to audit outputs and implementing deterministic security controls to prevent destructive actions."},{"start_ms":2180000,"title":"Risk in Non-Verifiable Domains","summary":"Analyzing the dangers of deploying agents in environments where success cannot be easily validated by unit tests or code compilers."},{"start_ms":2440000,"title":"The Challenges of Physical AI","summary":"Why robotic manipulation and tactile sensing present a much higher difficulty bar than digital agentic workflows."}],"topics":["AI Agents","Generative AI","Machine Learning","Autonomous Systems","Multi-Agent Systems","AI Safety","Computer Use Agents","Deep Learning"],"duration_seconds":3498,"processing_state":"processed","actions":[{"name":"request_transcript","method":"POST","url":"https://stenobird.com/v1/public/podcasts/dataframed/episodes/358-how-ai-agents-will-work-while-you-sleep-ruslan-salakhutdinov-professor-at-carnegie-mellon/transcription-requests","description":"Idempotently request low-priority transcript generation for this episode."},{"name":"read_markdown","method":"GET","url":"https://stenobird.com/podcast/dataframed/358-how-ai-agents-will-work-while-you-sleep-ruslan-salakhutdinov-professor-at-carnegie-mellon.md","description":"Read the agent-friendly Markdown representation of this episode resource."}]}}