Наредни састанак Семинара биће одржан онлајн у среду, 7. маја 2025. године, са почетком у 19 часова.
Предавач: Daniel Erasmus, CEO, Erasmus AI
Наслов предавања: FROM ARTIFICIAL TO SOCIAL INTELLIGENCE: CLIMATEGPT AND THE MOVE TOWARD HUMANIZING AI
Апстракт: Rather than positioning AI as an amorphous intelligence, as the cusp of AGI, this talk by Daniel Erasmus, founder of ClimateGPT, reframes AI models as complex but fundamentally instrumental systems—tools of conversation, memory, and inference. These models—“big statistical machines”—do not foreshadow the arrival of AGI but rather offer new ways to connect individuals, groups, and communities to knowledge, insight, and each other. In this light, AI becomes a medium for cultivating social intelligence: the emergent capacity to engage with large language models (LLMs) in ways that reshape communities of practice and support collaborative meaning-making. This form of social intelligence is critical in addressing a set of emergent global problems, and Daniel Erasmus argues that climate change is most pressing among these problems. Erasmus also argues that what matters is not just the sophistication of AI, but the timing and quality of its integration into shared human efforts—early action, early adaptation, and early mutual understanding are essential for navigating a future that may diverge radically from the speculative visions of „flying cars.“
Напомена: Регистрациона форма за учешће и линк за активно праћење предавања за регистроване кориснике (након логовања):
https://miteam.mi.sanu.ac.rs/asset/CW5nJWDSEZDj7p32p
Нерегистровани корисници могу да прате предавања на овом линку (без могућности активног учешћа):
https://miteam.mi.sanu.ac.rs/call/hR9vL94nD6QE8qQZj/xET9GcPMyR08nqH8lnS3SE7N5Vf00H7Lp9EBhsv6Lti