This OpenAI Forum session focused on how AI is accelerating research across disciplines—especially in the humanities and social sciences, where it’s enabling new ways to test theories, analyze culture, and train students in computational methods. The program framed AI as a powerful scientific tool that can compress long research timelines by handling tasks that may require deep, sustained reasoning and large-scale synthesis. Katherine Elkins shared “applied humanities” projects that model emotional arcs in novels, compare how translations reshape narrative patterns, and surface meaningful peaks and shifts that often align with what close-reading tends to notice. She also showed how students are using AI to explore cultural datasets—ranging from storytelling structures and social media dynamics to bias investigations in image generation, legislative text mining, and network/knowledge-graph analysis. Marco Uytiepo described how deep learning accelerates nanoscale brain-imaging analysis, turning months or years of manual reconstruction into days and helping researchers study circuit features linked to memory. Leonardo Impett argued that modern computer vision models don’t just analyze images—they embody a “machine visual culture,” and researchers can use art-historical methods to study both visual media and the cultural lens of the algorithms themselves. The event ended with a live Q&A where participants discussed responsible use with domain experts, creative uses of generative tools in storytelling, examples where AI changes research direction (not just speed), translation effects, long-term implications for analyzing AI-generated imagery, global archival preservation, and practical first steps for bringing AI methods into labs and classrooms.

OpenAI unveils grid-flexible data centers that cut peak load and lower costs, while enterprise AI adoption grows globally and boosts workers’ productivity.

The session explored how AI is reshaping education, beginning with OpenAI’s Olivia Pavco-Giaccia outlining why teaching and learning sit at the core of the company’s mission and how student adoption—now more than 40% of ChatGPT users worldwide—has accelerated AI’s integration into campuses, leading to large-scale deployments such as ChatGPT EDU across the CSU system and emerging research partnerships aimed at improving learning outcomes. She emphasized moving beyond fears of cheating to unlock personalized learning support, including early progress with Study Mode. Columbia University’s Moran Cerf, then connected AI to cutting-edge neuroscience and behavioral research, detailing how AI literacy is now essential for executives and students alike, and sharing striking findings from brain-interface studies that reveal how humans make decisions, process memories, and respond to engagement—insights he believes could transform pedagogy and human–AI collaboration. Miami Dade College’s Beth Muturi followed with a pragmatic model for “humanizing AI,” showing how community-embedded capstone projects allow students to build real AI solutions for local organizations, expanding opportunity and practical skill development. UC Berkeley’s Greg Neimeyer proposed a holistic AI pedagogy organized around three modes—minus AI, plus AI, and times AI—arguing that education must balance embodied human experience, critical engagement with AI, and transformational uses of intelligent systems to sustain meaning, collaboration, and truth in an AI-saturated era. Finally, UCLA’s Tina Austin presented her “Autumn Bloom’s” framework, a recursive, discipline-flexible alternative to Bloom’s taxonomy that shifts assessment away from grading AI-generated output and toward evaluating students’ comparative reasoning, critique, and metacognitive understanding, offering a path for assignments that integrate, challenge, or intentionally exclude AI depending on pedagogical need.