BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//hackathonmaxxing//EN
CALSCALE:GREGORIAN
X-WR-CALNAME:Hackathons in Berlin
BEGIN:VEVENT
UID:luma:evt-Yu7koJzotVJG3v9@hackathonmaxxing
DTSTAMP:20261122T163402Z
DTSTART:20261124T173000Z
DTEND:20261124T193000Z
SUMMARY:Recursive Self-Improvement: Superintelligence by 2028?
URL:https://lu.ma/v6am71hy?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=berlin
LOCATION:Hasso Plattner Institute (HPI)\, Potsdam\, Germany
DESCRIPTION:📍HPI main building\, D-Space (big space upstairs)\n\n🙋Open to everyone\, not just HPI students\n\nℹ️Sign-up helps us plan\, but is not required\n\nDue to the pressing topic about Recursive Self-improvement\, Julian Schulz and the HPI AI Safety Initiative decided to change the topic - We can still dive into questions about CoT-monitory\, mechanistic interpretability\, etc. in our discussion round after the main talk.\n\nAn AI that can autonomously research and train a smarter AI\, which trains a smarter one still\, which…\n\nOver the last month\, many AI safety researchers and forecasters have come to believe that this loop is imminent.\n\nWhat convinced them? Could the loop actually close? What happens if it does\, and can it be stopped?\n\nThis talk does not offer confident answers. Instead\, we go through the strongest evidence for and against: where AI systems stand today\, how much they already speed up AI research\, and what short timelines would imply for how we can help.\n\nJoin us in a talk and discussion with Julian Schulz! Whether you want to get a glimpse of what is going in AI Safety research or deep-dive into questions about RSI - everyone is welcome and all questions can be asked!\n\nJulian is an AI safety researcher at Principles of Intelligence\, applying mechanistic interpretability at scale\, and a Visiting Researcher at Meridian\, studying whether LLMs learn steganographic reasoning during RL. He teaches at the Iliad Intensive and ML4Good\, mentors at MARS and SAIGE\, and was a MATS scholar. Learn more about his path from physics and philosophy to AI safety: https://www.julianschulz.eu/\n\nBring your questions!\n\nBest\,HPI AI Safety Initiative (HAIS) https://lu.ma/v6am71hy?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=berlin
END:VEVENT
END:VCALENDAR
