The GitLab Foundation, in partnership with OpenAI, held its second annual demo day to showcase 15 innovative grantee projects aimed at driving economic mobility. The event highlighted how artificial intelligence can be used to improve the lives of individuals by increasing lifetime earnings and reducing economic barriers. Project presenters, in order of appearance: Terrina Govender Heath, Director, Open Weather Initiative Shweta Bhogale, Lead Researcher, Development Innovation Lab Caitlyn Brazill, President, Per Scholas Lois Lupica, Lead Designer, Community Economic Defense Project Paul Gilson, Vice President of Strategic Initiatives, Community Economic Defense Project Ayush Chopra, Co-Founder, Iceberg Systems Cassandra Okechukwu, Chief Science Officer, National Association of Community Health Centers Harpreet Marwah, Product Lead, TruePath Navigator, Career Path Services Allison Johnson, Director, Open UI, National Association of State Workforce Agencies Kelly Smyth, Chief, Office of Workforce Data & Research, New York State Department of Labor Ai-jen Poo, President, National Domestic Workers Alliance Elza Monteiro, Home Care Council Member, National Domestic Workers Alliance Christine Heitz, CEO, Colorado Thrives Steve Lee, CEO, SkillUp Coalition Deborah Singer, Chief Marketing Officer, Moms First Mike Marriner, Co-Founder & President, Roadtrip Nation Myles Sutholt, Head of Product and Design, Field Intelligence Kavitta Ghai, Co-Founder & CEO, Nectir

In this OpenAI Forum conversation, Sébastien Bubeck joins mathematicians Jacob Tsimerman, Daniel Litt, and Po-Shen Loh to explore how AI could reshape mathematical research and education. Filmed on August 11, 2026, the discussion examines what makes a mathematical result valuable, from solving long-standing problems to generating new ideas and discoveries. The panel also considers how students can use AI as a learning tool and, looking ahead, how AI could help mathematicians tackle harder questions and rethink mathematics’ role in the wider world.

This OpenAI Forum event offered a closer look at the Applied AI Engineering role at OpenAI and how the team turns powerful models into reliable, useful systems. Bonnie Chatterjee spoke with applied AI engineers Erika Kettleson and Thomas Li about their day-to-day work and collaboration with customers and teams across OpenAI. They explained how technical expertise and judgment shape everything from model performance to customer launches. They also talked about the importance of curiosity, ownership, and keeping up as the technology changes.

Richard Ho, OpenAI’s VP of Hardware, shares early results from Jalapeño, OpenAI’s first custom inference chip. He explains how it is already proven to make ChatGPT responses faster and Codex sessions and agents more responsive. He also discusses how OpenAI models helped the team design and improve the chip, and why developing the models, software, and infrastructure together was so important. Finally, Richard looks ahead to deploying Jalapeño across OpenAI’s compute infrastructure and building future generations of the chip.

Amy Deng shared how she used ChatGPT and Codex while navigating fatigue after treatment for a benign brain tumor. She described complex care as an information-management challenge, with important details spread across appointments, records, and daily experiences. AI helped her collect and organize that information so she could arrive at appointments better prepared and give her clinicians more context. She also demonstrated a Codex skill that turns spoken symptom notes into structured records when typing feels difficult. Her broader point was that AI can help patients participate more actively in their care while leaving medical decisions to clinicians. If you’d like to learn more about Amy’s work, visit amydeng.me.

OpenAI’s Chris Lehane and Morgan Dwyer sit down with the company’s head of finance communications, Josh King, for a brisk tour of the company’s latest updates. Lehane casts OpenAI’s momentum as a fast-turning flywheel: better models draw more users and grow more revenue. That, in turn, funds the compute needed for the next technological leap. Dwyer makes a compelling case, drawn from our paper, “Industrial Policy for the Intelligence Age,” that the road to superintelligence should be democratic, worker-centered and broadly beneficial, with outside researchers pressure-testing big ideas on economic security and frontier safety. Finally, Lehane fills us in on the new ChatGPT for teens and why it’s a safer, more thoughtful tool for learning.

OpenAI’s Chief Economist Ronnie Chatterji sat down with Columbia University economist Daniel Björkegren for a conversation about what AI could change for low- and middle-income countries, and what still has to happen for those gains to reach people. Daniel’s research paper followed about 500 teachers in Sierra Leone and more than 40,000 messages, showing why cost can make the web a poor fit in bandwidth-constrained settings. His research found that the average page used about 2.5 megabytes while an AI answer used only a few kilobytes, and found AI was over 700 times more data-efficient than loading a web page. In the paper’s latest update, AI was 98% cheaper, opening a wider discussion about local usefulness, trust, education, health, and new paths to growth. The conversation also marked Daniel’s selection as one of the first recipients of the OpenAI Economic Research Exchange, special recognition from a field of nearly 500 applications.

For families living with rare disease, the search for a diagnosis can stretch across years; this conversation made that uncertainty personal through Stav Rones’s story of finally finding a name for his condition. Researchers from Boston Children’s Hospital, Harvard, and OpenAI described how an AI-assisted workflow helped sift through genetic data and scientific literature, contributing to 18 diagnoses across 376 difficult cases. Just as importantly, they were clear that AI is not replacing clinicians: it narrows the field, presents the evidence, and leaves experts to test, interpret, and ultimately return answers to patients. The event offered a hopeful glimpse of a future in which more families can move sooner from uncertainty to understanding, community, and care.