$AMZN is developing an Alexa AI project codenamed Moonraker to handle multistep tasks like booking a ride and texting a friend. Internal docs reportedly project more than $100M in 2026 GPU costs making Moonraker the most expensive new Alexa+ initiative.

A project shaped by experimentation, trust and a genuinely wonderful collaboration. Thank you to the Kimi team for believing in the process and being such a pleasure to create with! #KIMI# #KIMIK3# #MOONSHOTAI# #K3# #BTS#

"Mooncake originated from a research collaboration between Kimi(Moonshot AI) and Tsinghua University. It was born from the need to solve the 'memory wall' in serving massive-scale models like Kimi K-Series. Since open-sourcing, it has evolved into a thriving community-driven project." GitHub:

One of my all time favorite collections is @moonbirds guided by one of the smartest leaders in the entire space @spencer . Definetly a lot of new shiny things to test drive in the NFT space recently but in the end I’ll always choose a legendary OG project like this one first. ❤️

With @inferact, @vllm_project is now officially verified by @Kimi_Moonshot to serve Kimi K3 with full accuracy, according to the Kimi Vendor Verifier. As the leading inference engine, vLLM delivers best results across vision, 1M context, and agentic coding; with prod quality.

Russia plans to build a nuclear power plant on the Moon Roscosmos says it aims to complete the project by 2036. After that, it plans to send nuclear-powered transport systems into space. Are they planning to build an oil refinery there too? Just in case.

After years of research, Project Suncatcher is scheduled to embark on its first test in orbit, launching a prototype satellite to evaluate how Google Tensor Processing Units (TPUs) perform in space. Learn more about the science behind this moonshot. ⬇️

Builders of the future 🤖 Our ARMADAS project is developing software and hardware that will autonomously assemble structures on future deep space missions, such as lunar infrastructure, landing pads, habitation components, and more. Learn how this work is preparing us for exploration on the Moon:

🧵AMA Recap｜Moonlight × UXLINK CEO Rolland Our AMA title was “Why UXLINK May Be One of the Most Mispriced Mass-Adoption Infrastructures in Web3?” In every Web3 cycle, “Mass Adoption” becomes the loudest catchphrase.  But after hosting countless AMAs and speaking with teams across ecosystems, I’ve learned one thing: very few projects are actually solving the hardest part of adoption, bringing real Web2 users into Web3 and keeping them there through real relationships, real usage, and real value. That’s exactly why I invited Rolland, CEO of UXLINK, to this AMA.  Our goal was simple: cut through the hype and examine who is quietly building the long-term growth infrastructure of Web3. What followed was a conversation that reframed how I, and likely many listeners, should think about mass adoption. Mass adoption is already happening, but not in the way most people think. Rolland began by acknowledging that mass adoption is no longer theoretical.  We’ve already seen projects like Catizen, CYBER, and PARTI drive explosive user growth in their respective domains, gaming, decentralized identity, and AI content.  Each of them is executing extremely well on a clearly defined track and has successfully pulled waves of new users toward Web3. But Rolland challenged us to look beneath the surface. While these projects shine within specific verticals, very few are building a persistent network of real people and real relationships.  UXLINK operates precisely in this overlooked layer, not as a spotlight application, but as the foundation beneath the ecosystem. From my perspective as the host, this was the first key insight: UXLINK is not competing for attention; it is competing to become indispensable. To explain this difference, Rolland introduced an analogy that stayed with me throughout the AMA.  He described most successful applications as “track leaders”, highly optimized products designed to win within a single scenario. UXLINK, by contrast, is building the “soil.” If other projects are digging wells on their own land, UXLINK is laying the underground water network. Instead of optimizing for one product form, UXLINK focuses on connecting real users, verifying social relationships, and creating a reusable growth layer that any project can build upon.  Tracks may change over time, but soil compounds. One of the most important moments of the AMA came when Rolland reframed UXLINK’s core mission.  Most Web3 projects ask, “How do we grow faster?” UXLINK asks a very different question: “How do we make the entire industry grow more easily?” That distinction explains why UXLINK doesn’t always look flashy during short-term market cycles. Infrastructure rarely does.  But once established, it becomes extremely difficult to replace.  From a host’s perspective, this also clarifies why UXLINK may be systematically undervalued, its value shows up in what others are able to launch, scale, and sustain because of it. Rolland then broke down UXLINK’s long-term value into four deep moats, and hearing them explained together made it clear why the project sits in a category of its own. First is the real social graph. In an industry filled with bots, scripts, and artificial activity, UXLINK insists on doing the hardest thing: connecting real people through acquaintance-based social networks.  This approach has enabled the genuine migration of tens of millions of Web2 users into Web3. Real relationships are an asset that cannot be fabricated or gamed. Second is OAOG, UXLINK’s cold-start engine. OAOG is not a marketing slogan but a precision-operated growth system.  By combining social trust, verifiable relationships, and fission mechanisms, it allows projects to bootstrap real communities across chains, regions, and markets, breaking the traditional cold-start curse in Web3.