In the Long Run

This episode explores the future of self-improving AI. MIT’s SEAL framework lets language models generate their own fine-tuning data and learning goals. Sakana AI’s Darwin Gödel Machine goes a step further: it rewrites its own code through evolutionary search, building a growing library of smarter agents. These advances point to a future where smaller, more adaptable models can keep learning on the fly.We also cover the conflict between OpenAI and the Google-backed audio startup iyO. OpenAI is investing $6.5 billion in Jony Ive’s hardware company (io) to develop a new AI companion. iyO is about to launch iyO One, AI-powered earbuds for voice-first interaction. The legal clash reflects rising expectations in AI-native hardware.Google DeepMind introduces Gemini Robotics On-Device. It runs locally on robots, removing the need for internet access and enabling real-time object control. The new SDK lets developers fine-tune models with just 50 demonstrations, making it quicker to adapt to new tasks.Your Summer Survival Kit: listen again to our episode “Attention Is All You Need”, read Klara and the Sun, and let Jim explore AI tools like Synthesia, Canva AI, Notion AI and Sana Agents (so you don’t have to). Instead, take the time to reflect on how AI is shaping your field and get ready for what autumn brings.---Papers“Self-Adapting Language Models (SEAL)” – Zweiger et al., MIT, 2025“The Darwin Gödel Machine” – Sakana AI & UBC, May 2025ClipsSam Altman + Jony Ive IO announcement (YouTube)iyO earbud reveal (TED talk)Generalist robotics demoGemini Robotics SDK launchSummer Reading & ListeningKlara and the Sun – Kazuo IshiguroAI och makten över besluten – Steinrud et al.BBC In Our Time (podcast)Hard Fork (New York Times)The Ezra Klein ShowTools to TestSynthesia (AI Video)Canva AINotion AISANA agents

In this episode of In the Long Run, hosts Marie, Joao, and Jim take a good hard look at the complexities of human identity in an increasingly AI-driven world. They discuss WorldCoin, a project co-founded by Sam Altman, which aims to verify human identity through biometric data using “orbs” to differentiate humans from AI bots. The conversation explores the challenges of trust and identity verification in the age of deepfakes and AI agents. They debate whether it’s essential to distinguish between humans and AI, and how concepts like WorldCoin could impact privacy, security, and societal norms. The hosts also examine the role of governments versus private entities in managing identity verification and the potential implications of decentralized systems. They touch upon the distribution of benefits from Artificial General Intelligence (AGI) and the ethical considerations of such advancements. Join us as we navigate the intricate intersection of technology, trust, and what it truly means to be human in the digital age.

In this episode of In the Long Run, Joao, Marie and Jim dive into the exciting world of OpenAI’s latest release—the O1 preview model. Designed to spend more time “thinking” before responding, this new AI model is breaking barriers in complex problem-solving, coding, and mathematics. The hosts discuss how this represents a significant leap in AI capabilities, possibly even challenging our definitions of intelligence. They also explore the evolving relationship between humans and AI, debating whether AI could be considered a new form of intelligence or even a separate “species.” The conversation touches on the importance of collaboration between humans and AI, the ethical considerations of AI-assisted creations, and how organizations can adapt to these rapid technological advancements. Join us as we navigate the implications of this paradigm shift and ponder what it means for the future of work, research, and society at large.