Loft Orbital to test AI models on spacecraft for Earth observation

spacenews.com Jun 23, 2026

Loft Orbital to test AI models on spacecraft for Earth observation

Curated by Shubham · TechCapsules · Jun 23, 2026

AI-summarised brief · reviewed before publication

Loft Orbital is working with NASA's Jet Propulsion Laboratory to test artificial intelligence on spacecraft for Earth science monitoring. The project, called Federated Autonomous Measurement, aims to automate the tip-and-cue process, where one spacecraft identifies features for another to observe in detail. Tests started this month using a Loft Orbital spacecraft with JPL's AI software, with additional tests planned for 2027 and 2028. The goal is to enable real-time processing and analysis of data without downlinking large amounts of data. This approach uses AI models trained on a large corpus of data to recognize features without being told what to look for.

💡 Why It Matters

· Real-time Earth observation enabled by AI on spacecraft can quickly detect and respond to natural disasters like wildfires and marine pollution events.
· Automated tip-and-cue processing also enhances security, military, and intelligence applications by providing rapid information to trigger actions.

Read full article on spacenews.com

Mitigating vendor lock-in with Sakana AI Fugu multi-agent models

Sakana AI launched Fugu, a platform that orchestrates multi-agent operations to mitigate single-vendor dependency risks in enterprise deployments. Fugu uses an orchestration language model that calls upon a pool of varied models to complete tasks, routing queries internally and deciding whether to resolve prompts directly or assemble a team of expert models. The system handles model selection, delegation, verification, and synthesis internally, providing a resilient architecture for AI sovereignty. Fugu targets geopolitical and regulatory risks associated with AI sourcing, and its swappable agent pool allows for dynamic routing around restricted providers. Two deployment tiers are available, with Fugu Ultra targeting complex analytical problems.

💡 By providing a hedge against sudden supply chain disruptions, Fugu enables companies to maintain service continuity despite export controls or vendor restrictions. Fugu's ability to outperform monolithic models in defect detection and security vulnerability identification makes it a valuable tool for software development and cybersecurity teams.

thenewstack.io · Jun 22

“An agent is an LLM and a harness”: What Nvidia really thinks…

Nvidia's Director of Deep Learning Software, Bryan Catanzaro, has shed light on the company's stance regarding OpenClaw, a new open-source project that aims to create a more efficient and scalable deep learning framework. Catanzaro described OpenClaw as "an agent is an LLM and a harness," suggesting a symbiotic relationship between the framework and large language models (LLMs). This statement implies that OpenClaw is designed to work seamlessly with LLMs, potentially unlocking new levels of performance and efficiency in deep learning applications.

💡 Nvidia's endorsement of OpenClaw highlights the growing importance of open-source collaboration in the deep learning space. By embracing OpenClaw, Nvidia is signaling its commitment to advancing the field through community-driven innovation, rather than relying on proprietary solutions. This shift could have significant implications for the development of future AI applications, as open-source frameworks like OpenClaw become increasingly influential in shaping the direction of the industry.

timesofindia.indiatimes.com · Jun 21

3,000 Anthropic employees have a question after US government ‘banned’ company’s most…

Anthropic's 3,000 employees are left wondering why the US government suddenly shut down the company's most powerful AI models, Fable 5 and Mythos 5, with less than 90 minutes' notice. The order cited national security concerns, but no explanation was provided, leaving managers to warn customers about an outage without knowing the reason. The models were pulled mid-task, and employees are speculating about the cause, with some suspecting it may be personal rather than technical. The company had to figure out the trigger on its own, tracing it to a research paper from Amazon, which showed Fable 5 could identify flaws in vulnerable code.

💡 The sudden shutdown raises questions about government oversight of AI development, highlighting the lack of transparency and communication between regulators and tech companies. Government intervention in AI development can have far-reaching consequences for the industry.

techcrunch.com · Jun 21

In the Weights is your new AI-centric vanity search

Thomas Dimson and Joey Flynn have created In the Weights, a website that measures how well AI models recall individuals without using web search. The site queries different models with a question about a person's name and assigns a strength score based on the results. The score is calculated by clustering similar descriptions together, with higher scores indicating a stronger presence in the AI models. Users can see which models returned which answers and potential hallucinations. The site has gained popularity, with users comparing their scores and exploring the results. The creators plan to further analyze the results and models. The site's popularity has been significant so far, with many users drawn to its unique concept and retro design.

💡 In the Weights taps into people's desire for digital legacy and immortality, sparking curiosity about how AI models perceive and remember individuals. The site's comparison factor and easy-to-understand scores have made it a fascinating tool for self-discovery and exploration.

techcrunch.com · Jun 20

From PGP to Mythos: a brief history of export controls that didn’t…

The White House ordered Anthropicto restrict the export of its AI models Fable and Mythos to non-US entities and foreign nationals, citing national security concerns. The move came after Anthropic gave a South Korean telecom access to Mythos, sparking US officials' alarm over suspected ties to China, and Amazon's researchers found a way around Fable 5's safeguards. As a result, Anthropic pulled the plug on both models, which had limited access to around 150 vetted companies and government organizations. This episode tests the US government's ability to use export controls to contain frontier AI, with significant implications for Anthropic and other AI labs. The US has a mixed track record in using export controls to limit cyber technology proliferation.

💡 The outcome of this standoff will shape the rulebook for other AI labs, determining their access to foreign markets. Export controls' effectiveness in containing AI will influence the development of future AI technologies.

asiatimes.com · Jun 20

Hidden in plain sight: the AI within Xiaomi

A previously unknown AI model from Xiaomi recently appeared on a developer platform, reaching the top of the usage charts and serving over one trillion tokens before its origin was discovered. This counterexample challenges the notion that building a cutting-edge AI model requires a long head start and significant resources. Xiaomi's model is notable, given the company's diverse product portfolio and significant revenue. The model's success is particularly striking, as Xiaomi only began building its large-model team in early 2023 and shipped its first language models that same year. The company's progress in AI development has been rapid, with its model now considered more exciting than some pure-play AI companies.

💡 Xiaomi's achievement undermines the assumption that a decade-long head start is necessary for building a competitive AI model. Its success also raises questions about the valuation of companies that make AI models compared to those with more diverse product offerings.