BEGIN:VCALENDAR VERSION:2.0 PRODID:Data::ICal 0.24 BEGIN:VEVENT DESCRIPTION: 'Title: Evil Digital Twin: Learn Psychological Manipulation with an\n Uncensored LLM\n When: Saturday\, Aug 12\, 11:00 - 11:59 PDT \n Where: Caesars Forum - Summit - 224 - Misinfo Village - [1]Map\n Sp eakers:Ben D. Sawyer\,Matthew Canham\n\n SpeakerBio:Ben D. Sawyer\n Dr . Ben D. Sawyer is an applied neuroscientist and human factors\n enginee r known for using brainwaves\, eye movements\, and mathematical\n theory to build better human-machine teams. His models and algorithms\n power trustworthy machines that work with their human partners. His\n design r ecommendations are leveraged by Fortune 500 companies. His\n work has be en covered by Forbes\, Reuters\, Fast Company\, and The BBC\,\n and more . Dr. Sawyer's postdoctoral work at MIT was in collaboration\n with indu stry including Google\, Jaguar-Landrover\, Honda\, DENSO\,\n Monotype\, and Panasonic. A two-time Repperger Research Fellow with the\n Air Force Research Laboratory (AFRL)\, he performed research with the\n 711th Hum an Performance Wing in both their Applied Neuroscience and\n Battlefield Acoustics (BATMAN group) divisions. He is a recipient of\n The Human Fa ctors Prize\, for Cybersecurity research\, The K.U. Smith\n Award\, for consumer electronics work investigating driving distraction\n and Google Glass\, and an Outstanding Dissertation Award for work\n investigating the applied psychophysics of warfighter multitasking.\n Dr. Sawyer is pr esently faculty in Industrial Engineering and\n Management Systems and t he Institute for Simulation and Training at\n UCF\, and the Director of LabX\, an applied neuroscience group\n addressing human performance. In addition to academic pursuits\, Dr.\n Sawyer provides consulting and exp ert witness services in matters\n including distraction\, attention\, an d error. He is the co-founder of\n Awayr\, an AI startup within MIT's Th e Engine. In his leisure time\, Ben\n enjoys adventure travel with his w ife\, swimming\, sailing the Charles\,\n and building things. He volunte ers\, helping people to learn to repair\n their own things. He does not enjoy writing about himself in the third\n person and will now stop.\n\n SpeakerBio:Matthew Canham \, CEO at Beyond Layer 7\n Dr. Matthew Canh am is the CEO of Beyond Layer 7\, a security consulting\n firm specializ ing in mitigating insider risk\, human-centric security\n solutions\, an d data analytics. He is a founding member of the\n Cognitive Security In stitute\, a non-profit organization dedicated to\n research and understa nding the nuances of defending human and\n artificial decision makers fr om cognitive attacks. He currently holds\n an affiliated faculty member appointment with George Mason University\,\n where the focus of his rese arch orients around synthetic media\n –deepfake– social engineering attacks\, cognitive security\, and\n insider risk. Dr. Canham was previo usly a Special Agent for the FBI\,\n where he investigated cybercrimes\, insider threats\, and ran cyber\n HUMINT operations as a certified onli ne covert employee. During his\n tenure with the FBI\, he also served as a subject matter expert with\n the Behavioral Analysis Program (BAP) an d as the Technical Liaison\n between the Operational Technology Division (OTD) and the Cyber\n Behavioral Analysis Center (CBAC). Dr. Canham has previously taught\n courses on synthetic media attacks\, cognitive warf are\, human-centric\n security\, and insider risk for\; NATO\, NASA (Ken nedy Space Center)\,\n DARPA\, MIT\, BI (Oslo\, Norway)\, Aberdeen Provi ng Ground\, ISC2\, and the\n FBI. He holds a PhD in Cognitive Psychology with special expertise in\n human centric security\, data analytics\, a nd behavioral engineering\; as\n well as GIAC certifications in Certifie d Forensics Examiner (GCFE)\,\n Certified Auditing Wireless Networks (GW AN)\, Mobile Device Security\n Analyst (GMOB)\, and Certified Incident H andler (GCIH).\n\n Description:\n This workshop will delve into the un charted territory of Uncensored\n Large Language Models (LLMs) and explo re their latent potential. This\n immersive session invites cybersecurit y experts\, researchers\, and\n enthusiasts alike to understand the prof ound capabilities of\n uncensored LLMs displayed in misinformation and m anipulation tactics\,\n and how malicious actors might exploit this capa bility.\n\n During the workshop\, attendees will gain a practical unders tanding of\n how LLMs can deploy sophisticated strategies pulled from ps ychological\n literature and advertising science. From harnessing cognit ive biases\n and social norms to capitalizing on habit loops\, the mecha nisms of\n manipulation are as intriguing as they are unsettling. An unc ensored\n LLM\, ripe for exploration\, serves as our playground\, provid ing\n invaluable insights into its interaction patterns.\n\n But it's not just about understanding these models\; it's about looking\n beyond the horizon. Participants will also be introduced to the\n concept of 's hadow prompts\,' concealed instructions that a compromised\n or maliciou s LLM runs alongside standard user prompts. Unraveling the\n intricacies of these shadow prompts will reveal how they can\n stealthily shift the outcomes of LLM interactions.\n\n This workshop is more than just a one -off event. All participants will\n automatically become members of our burgeoning ""Evil Digital Twin""\n community. This membership comes with the opportunity to engage in an\n ongoing learning journey\, unveiling the less-discussed capabilities of\n LLMs. By fostering a collaborative environment\, we aim to empower\n members to not only understand these e volving digital entities but\n also devise robust defenses against them within their respective\n organizations and institutions.\n\n Two deep experts in psychology and cybersecurity and the intelligence\n communit y will be your guides.\n\n '\n\n 1. #CaesarsSummitBR\n\n\n DTEND:20230812T185900Z DTSTART:20230812T180000Z LOCATION:MIV - Caesars Forum - Summit - 224 - Misinfo Village SUMMARY:Evil Digital Twin: Learn Psychological Manipulation with an Uncenso red LLM END:VEVENT END:VCALENDAR