The Rundown AI is reporting that OpenAI CEO Sam Altman has proposed a U.S.-led forum to establish artificial intelligence safety standards and determine access to advanced AI models. It said Altman made this call in an op-ed following June's G7 summit in France, where he discussed AI regulation with government leaders. The Rundown AI also reported that OpenAI has discussed offering the U.S. government a 5% stake in the company, suggesting other U.S. AI labs could contribute to a dividend fund for wealth redistribution. Altman emphasized that "democratic institutions must not cede their responsibilities to AI labs" and that "citizens and their elected representatives must make the rules."
The Rundown AI is also reporting on the potential for smart glasses to replace smartphones, noting that Meta CEO Mark Zuckerberg has long championed this idea. It said Meta currently holds an estimated 80% of the global smart glasses market. The publication highlighted features of the new Meta Glasses, such as GPS, live translation, music, voice notes, and a built-in camera, which it said offer advantages over phones for certain tasks like sightseeing and hands-free cooking. However, it noted that battery life and limited app integration currently prevent them from fully replacing smartphones.
In other artificial intelligence news, The Rundown AI said research from Mira Murati's Thinking Machines Lab and Bridgewater demonstrates the effectiveness of specialized AI models. It reported that a small custom AI, trained on Bridgewater's investment judgment, outperformed leading frontier models like GPT, Claude, and Gemini variants on news filtering investment tasks. This specialized model achieved 84.7% accuracy at 13.8 times less cost than frontier models, which averaged around 50% accuracy. The Rundown AI highlighted this as evidence that companies may not always need comprehensive models but rather highly specialized, cost-effective solutions for specific tasks.
The Rundown AI also noted that Palantir CEO Alex Karp criticized frontier AI giants on CNBC, stating that enterprises "are paying for tokens that create no value" and that labs "are stealing weights and alpha." It said Microsoft launched "Frontier Company," committing $2.5 billion and 6,000 in-house engineers to help clients build and operate AI systems. Additionally, China's Kling AI secured $2 billion in funding, aiming for global expansion after OpenAI's Sora shutdown. Anthropic reportedly approached Samsung about manufacturing its own AI chip.