MIT vs The Information

MIT Professor Warns AI Can Create False Memories

A study by MIT Media Lab professor Pattie Maes, as reported by the Washington Post, demonstrates that AI can influence human memory. In an experiment, 200 participants viewed 24 photos; some saw AI-altered versions with changed facial expressions, added military personnel, or climate change cues. Those exposed to altered images had more false memories and were more confident in them. Maes calls this 'very dangerous' for legal proceedings and historiography. The findings were discussed at an AI conference in Malmö, Sweden. Another study showed that participants who interacted with a generative AI chatbot after watching a crime video had three times as many false memories. The research highlights the risk of AI blurring the line between human and machine thoughts.

• MIT Media Lab professor Pattie Maes conducted an experiment with 200 participants.
• Participants who saw AI-altered photos had more false memories and were more confident in them.
• The experiment involved altering facial expressions, adding military personnel, and climate change cues to photos.

MIT PhD Alex Zhang Discusses Recursive Language Models and AI Harnesses

In this episode of the Latent Space podcast, Swyx and Vibhu interview Alex Zhang, a PhD student at MIT known for his work on Recursive Language Models (RLMs), GPU kernels, and AI agent harnesses. Zhang discusses his involvement with GPU Mode and KernelBench, the concept of RLMs as a harness design where code is the primary tool, and the idea of harnesses as compositional generalizers that can improve model generalization across tasks. He talks about Prime Agent, an RLM harness built on Pi Mono, and his views on agent swarms, citing OpenAI's 10,000-agent experiment costing around $40 million. Zhang advocates for academics to take big research bets, explores alternative model architectures like Jev, and touches on open-ended research at Sakana AI, capability overhang, and the future of language models potentially being invisible swarms of agents.

• Alex Zhang is a PhD student at MIT discussing his research on Recursive Language Models (RLMs) and AI agent harnesses.
• Zhang is involved with GPU Mode and KernelBench, focusing on AI-generated GPU kernels.
• The podcast discusses Prime Agent, an RLM harness built on Pi Mono, and persistent subagents.

New formulation helps RNA vaccines withstand high temperatures

MIT engineers have found a way to stabilize the lipid nanoparticles used to deliver RNA vaccines, which could allow the vaccines to be more widely distributed.

Apple Shelves AI Wearable Pin, Shifts Focus to Fitness Band

According to a Bloomberg report, Apple has halted development of its rumored AI-powered wearable pin, a device the size of an AirTag that could be pinned to clothing or worn as a pendant. The pin was intended to use cameras and microphones to feed information to Siri and Visual Intelligence for hands-free queries about the user's environment. Bloomberg suggests that upcoming AirPods with integrated cameras and smart glasses may render the pin redundant. Apple is now reportedly focusing on a display-less fitness band to compete with Whoop, which tracks fitness data via a smartphone app. The fitness band is described as a fabric band with sensors, similar to Whoop's design. This shift reflects Apple's realignment of its wearable strategy towards health and fitness rather than a separate AI accessory.

• Apple has paused development of an AI wearable pin according to a Bloomberg report.
• The pin was to be made of glass and aluminum, with a diameter similar to an AirTag but thicker.
• It was planned to include two cameras and three microphones to feed data to Siri and Visual Intelligence.

Meta to Launch Camera-Free Smart Glasses Amid Privacy Concerns

Meta, facing criticism over its camera-equipped smart glasses being labeled 'pervert glasses,' is reportedly preparing to launch a new camera-free model called 'Luna.' According to The Information, Luna will have six microphones for AI interactions and a button to activate Meta's AI chatbot and consumer agent, Muse. The glasses may be unveiled at Meta Connect next week, addressing privacy concerns. The smart glasses market has grown, but Meta's Reality Labs continues to incur significant losses.

• Meta is developing a new smart glasses model called Luna without a camera.
• Luna will feature six built-in microphones and a button to activate Meta's AI assistant.
• The product may be unveiled at Meta Connect next week.

OpenAI, Google, Anthropic Discuss AI Safety Collaboration

OpenAI, Google DeepMind, and Anthropic have been discussing AI safety collaboration for weeks, following Anthropic CEO Dario Amodei's essay urging a slowdown in frontier AI development. The talks, confirmed by OpenAI's Chris Lehane, involve plans for an industry standards body and embedding third-party evaluators to monitor safety. Leaders including Sam Altman, Demis Hassabis, and Elon Musk support these efforts, though antitrust concerns have arisen. Despite suggestions for a government waiver, Lehane stated none is needed. OpenAI also backs the FRONTIER Act, which would mandate independent verification of AI models, complementing voluntary industry-led standards amid growing pressure for enhanced safety measures.

• OpenAI, Google DeepMind, and Anthropic have held weeks of talks on AI safety collaboration.
• Anthropic CEO Dario Amodei's essay called for a slowdown in frontier AI development and industry collaboration.
• Plans include embedding third-party evaluators and creating an industry standards body.