In collaboration with Mach33,ARKInvest open-sourced its #SpaceX model, the only open-source model in the market, if I am not mistaken, which should help put into perspective the massive opportunities, even before solar-powered data centers, in space.Sam Korus (@skorusARK):If for some reason you're randomly wondering what SpaceX might be worth...here is ARK's open source model https://t.co/wOhzAtISSt ...

AI Ecosystem & Investment - Nvidia's CEO expresses regret for not investing more in AI companies, viewing AI as encompassing energy, chips, models, and applications, highlighting Nvidia's role in building the AI infrastructure [1] - Reflections, an open-source AI model startup, secures $2 billion in Series B funding [3] - Nvidia is financing AI startups, enabling them to use the funds to access Nvidia's chips, indicating a circular investment model [6] Open Intelligence & Competition - Reflections aims to build a "frontier open intelligence" accessible to all, contrasting with the concentration of AI control in a few closed companies [4][10] - Open intelligence differs from open source in that it requires significant capital to train models, necessitating a commercial business model [12] - Reflections' mission is to bring the open-weight frontier back to the US, competing with open models from China, particularly in light of DeepSeek's launch in December 2024 [14] Market Strategy & Future - Reflections positions itself to serve large enterprise and sovereign entities, moving "up the Maslo's hierarchy of AI markets" [18] - The company aims to be the "last generational company" in the AI paradigm, focusing on open intelligence for enterprise solutions [18] - Reflections emphasizes being a reliable and compliant provider for enterprises seeking open intelligence, particularly given concerns about the data provenance of some models [16][17] Revenue Growth & Margins - The AI industry is experiencing unprecedented revenue growth from both foundation model makers and companies building on top of them [7] - While inference can be optimized for good margins, continued investment in training next-generation models is driving expenses [9]