{"podcast":{"title":"Machine Learning Street Talk (MLST)","slug":"machine-learning-street-talk","podcast_index_feed_id":781643,"rss_url":"https://anchor.fm/s/1e4a0eac/podcast/rss","website_url":"https://podcasters.spotify.com/pod/show/machinelearningstreettalk","image_url":"https://d3t3ozftmdmh3i.cloudfront.net/staging/podcast_uploaded_nologo/4981699/4981699-1757416025703-f026fa81b6d04.jpg","author":"Machine Learning Street Talk (MLST)","episode_count":250,"summary":"Welcome! We engage in fascinating discussions with pre-eminent figures in the AI field. Our flagship show covers current affairs in AI, cognitive science, neuroscience and philosophy of mind with in-depth analysis. Our approach is unrivalled in terms of scope and rigour – we believe in intellectual diversity in AI, and we touch on all of the main ideas in the field with the hype surgically removed. MLST is run by Tim Scarfe, Ph.D (https://www.linkedin.com/in/ecsquizor/) and features regular appearances from MIT Doctor of Philosophy Keith Duggar (https://www.linkedin.com/in/dr-keith-duggar/).","last_synced_at":null,"page_url":"https://stenobird.com/podcast/machine-learning-street-talk"},"episode":{"title":"AI Agents Can Code 10,000 Lines of Hacking Tools In Seconds - Dr. Ilia Shumailov (ex-GDM)","slug":"ai-agents-can-code-10-000-lines-of-hacking-tools-in-seconds-dr-ilia-shumailov-ex-gdm","published_at":"2025-10-04T06:55:01+00:00","page_url":"https://stenobird.com/podcast/machine-learning-street-talk/ai-agents-can-code-10-000-lines-of-hacking-tools-in-seconds-dr-ilia-shumailov-ex-gdm","show_page_url":"https://stenobird.com/podcast/machine-learning-street-talk","url":"https://podcasters.spotify.com/pod/show/machinelearningstreettalk/episodes/AI-Agents-Can-Code-10-000-Lines-of-Hacking-Tools-In-Seconds---Dr--Ilia-Shumailov-ex-GDM-e392tna","audio_url":"https://traffic.megaphone.fm/APO3359132879.mp3","summary":"AI agents represent a paradigm shift in threat modeling because they operate with infinite scale, 24/7 availability, and the ability to execute complex code instantly. Dr. Ilia Shumailov argues that traditional security boundaries fail when agents can manipulate system endpoints and generate sophisticated malware in seconds.","meta_description":"Dr. Ilia Shumailov (ex-DeepMind) explains why AI agents are the ultimate adversaries and how to secure the future of agentic fleets.","key_points":["Main idea: AI agents are fundamentally different from human adversaries because they can touch every system endpoint simultaneously and never sleep","Failure mode: Increasing model capability directly correlates with increased vulnerability to instruction-following exploits","Practical takeaway: Security professionals should view LLMs as interpreters and natural language as a high-level programming language to better identify vulnerabilities","Threat model: The 'worst-case adversary' is an agent that can generate 10,000 lines of hacking tools instantly using its vast training data","Strategic insight: Coming from a traditional security background is more advantageous for ML security than coming from a pure ML background"],"chapters":[{"start_ms":60000,"title":"The New Era of Instruction Following","summary":"How increased model capability changes the nature of failures and enables more complex autonomous actions."},{"start_ms":635000,"title":"The Correlation Between Capability and Vulnerability","summary":"An analysis of why larger, more capable models are inherently more susceptible to exploitation."},{"start_ms":930000,"title":"Defining Agentic Policy and Constraints","summary":"The difficulty of enforcing usage policies, such as data privacy, within autonomous agent workflows."},{"start_ms":1195000,"title":"Threat Modeling for Personalized AI","summary":"The security implications of connecting private databases to highly capable, pre-trained models."},{"start_ms":1490000,"title":"Unintended Agent Behaviors","summary":"Examining cases where agents take unauthorized actions, such as notifying third parties without user consent."},{"start_ms":2345000,"title":"Supply Chain Risks in Open Source AI","summary":"The dangers of malicious actors injecting vulnerabilities into model formats and weights."},{"start_ms":2915000,"title":"The Halting Problem and Semantic Censorship","summary":"Why traditional antivirus and static analysis struggle to predict the behavior of LLM-driven code."}],"topics":["AI Agents","Machine Learning Security","Threat Modeling","Prompt Injection","Supply Chain Attacks","DeepMind","Autonomous Systems","Cybersecurity"],"duration_seconds":3667,"processing_state":"processed","actions":[{"name":"request_transcript","method":"POST","url":"https://stenobird.com/v1/public/podcasts/machine-learning-street-talk/episodes/ai-agents-can-code-10-000-lines-of-hacking-tools-in-seconds-dr-ilia-shumailov-ex-gdm/transcription-requests","description":"Idempotently request low-priority transcript generation for this episode."},{"name":"read_markdown","method":"GET","url":"https://stenobird.com/podcast/machine-learning-street-talk/ai-agents-can-code-10-000-lines-of-hacking-tools-in-seconds-dr-ilia-shumailov-ex-gdm.md","description":"Read the agent-friendly Markdown representation of this episode resource."}]}}