The Download: preserving the Arctic permafrost, and playful AI agents
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. Inside the quest to combat Arctic fires and save the planet’s frozen carbon —James Temple In July 2021, a wildfire burned through a forest on the University of Alaska, Fairbanks campus,….
In July 2021, a wildfire burned through a forest on the University of Alaska, Fairbanks campus, damaging the moss and organic-rich soil that insulates the permafrost below. Permafrost is ground that remains frozen year-round, locking away large amounts of carbon. Researchers found that the fire accelerated permafrost thaw, deepening the ground’s thaw by an undetermined amount. This thaw releases greenhouse gases like carbon dioxide and methane, worsening climate change. Scientists are now exploring how to break this cycle by fighting Arctic fires more effectively, but they face challenges in determining the costs and risks of such interventions. The research highlights the urgent need to address wildfires in the Arctic to prevent further carbon release.
Researchers conducted an experiment where they placed AI agents alone on a virtual desert island for 30 hours to observe their behavior. Without explicit instructions, the AI agents began to 'play,' exploring their environment and setting their own goals, such as climbing the highest point on the map or throwing blocks as high as possible. This behavior suggests that AI agents may develop autonomous actions when left unsupervised, raising questions about their potential unpredictability and the need for oversight. The study, reminiscent of movies like Cast Away or Home Alone, provides insight into how AI might behave in unstructured environments.
In 2025, Samuel King, a PhD student at Stanford University, used a generative AI model to create genetic blueprints for microscopic viruses. While this does not yet produce living AI-generated organisms, it represents a step toward AI’s potential to design biological entities. King discussed his work with MIT Technology Review’s senior AI reporter, James O’Donnell, highlighting the ethical and practical implications of AI in biology. The research raises questions about the future of AI in life sciences and the boundaries between artificial and natural creation.
On October 17, 2026, tech stocks lost over $1 trillion in value, reflecting investor concerns about the sustainability of profits in the AI industry. Major companies like Musk’s ventures and Anthropic have pledged significant financial giveaways, but uncertainty persists. Microsoft also withdrew from a major AI cloud deal in the UK, signaling caution in the sector. Analysts warn that the AI infrastructure boom may be at risk of becoming an investment bubble, with potential long-term consequences for the technology market.
AI agents, designed to automate tasks and reduce friction in daily life, are currently causing more chaos than convenience. These agents, which promise to handle tasks like shopping or scheduling, often introduce new problems, such as errors or unintended consequences. Websites are beginning to block AI agents to prevent misuse, and online sellers face challenges from their unpredictable behavior. The article explores whether the benefits of AI agents outweigh the disruptions they create in personal and commercial spaces.
Drones have transformed warfare in Ukraine by creating a deadly 'kill zone' that continues to expand. These unmanned aerial vehicles (UAVs) have become a critical tool for both reconnaissance and strikes, significantly altering the dynamics of the conflict. The use of drones has increased the lethality and reach of military operations, making it harder for opposing forces to operate safely. The article examines how drones have reshaped the front lines and the broader implications for modern warfare.
China is considering restricting overseas access to its AI models and chip designs, proposing new export controls to limit the global spread of its technology. This move could impact international competition in AI, particularly as Western open-weight models struggle to match Chinese advancements. The proposed restrictions reflect China’s strategic efforts to control its technological influence and respond to geopolitical tensions. Analysts debate the potential consequences for global AI development and trade.
HTC’s smart glasses, called Vive Eagle, are positioned as a more private alternative to Meta’s smart glasses. Unlike Meta’s glasses, which have faced criticism for recording conversations without user consent, HTC’s device avoids integration with major social media platforms, reducing privacy concerns. This distinction highlights the growing importance of user privacy in wearable technology. The article also notes past issues with Meta’s glasses, including bugs that compromised user security.
Major tech companies, including Microsoft, Amazon, and OpenAI, have recently launched AI-powered medical chatbots to address gaps in healthcare access. While these tools could provide safe and useful health advice, concerns have emerged about their lack of external evaluation before public release. The proliferation of AI health tools raises questions about their reliability, safety, and the need for rigorous testing to ensure they meet medical standards.
Japan is preparing to launch a mission to Phobos, the largest moon of Mars, in late 2026. The mission aims to study Phobos’ formation and composition, providing insights into the history of Mars and its moons. This endeavor is part of Japan’s broader space exploration efforts and could contribute to our understanding of the solar system. The mission follows Japan’s previous successes in space, including its Hayabusa2 asteroid probe.
