ASI Researchers
60 posts
@ASI_Researchers
MeTTa, GSLTs, agentic architectures & alignment. Open-source theory 🤝 scalable AI practice. For researchers building a safer superintelligence @asi_alliance
We are pioneering research and infrastructure to achieve decentralized human-level AGI.
On the latest episode of The Ten Reckonings of AGI series, our CEO, Dr. @bengoertzel, sat down with technologist Jaron Lanier to explore the practical realities of building and governing advanced systems, and debate how far our empathy should extend to AI and future AGI.
While LLMs and other contemporary neural network architectures remain insufficient for achieving human-level AGI due to a lack of internal abstraction, which limits their ability to make creative leaps too far beyond their training data, they nevertheless demonstrate impressive practical capabilities across a wide range of applications. These capabilities can be replicated via non-neural architectures; however, there is a strong pragmatic motivation to keep what works and integrate it with other components that operate differently. This militates toward hybrid AGI architectures, e.g., neural-symbolic and neural-symbolic-evolutionary. In broad terms, current neural networks excel at pattern recognition, handling ambiguity, and learning from examples. Symbolic systems currently shine at explicit reasoning, structured manipulation, and explanatory transparency. Running them in isolation means constantly shuttling data back and forth, losing context in translation, and missing opportunities for mutual enhancement. Hyperon eliminates these barriers by making neural and symbolic components first-class citizens in the same computational space. Features, rules, proofs, options, and activations all become Atoms in shared memory, operating under a unified scheduling model. This means reasoning can directly guide where neural networks focus their attention, while neural networks can propose structured hypotheses back to the reasoning system–all without serialization boundaries or synchronization bottlenecks.
Yann LeCun says there is no such thing as general intelligence Human intelligence is super-specialized for the physical world, and our feeling of generality is an illusion We only seem general because we can't imagine the problems we're blind to "the concept is complete BS"
Achieving human-level AGI through Hyperon would not be a terminus but rather a beginning of a new phase of development, as the practical, creative, and analytical capabilities obtained increasingly turned themselves to improvement of the AGI system’s own capabilities.
As Title Sponsor for @HackIndiaXYZ 2025, we strengthened our presence across India’s academic institutions and developer communities, advancing the nation’s AI and Web3 innovation ecosystem. This nationwide initiative brought together over 29,700 students across 31 regional hackathons, making it the largest gathering of Web3 and AI developers in the country. The series culminated in the National Finals on September 26 at The Leela Ambience Convention Hotel in Delhi, with the presence of our CEO, Dr. @bengoertzel, and Angelica @NefertitiStrong, Founder of @BeyondtheCodeAI, where top innovators and the most promising projects from across the nation came together. The hackathon comprised 31 regional events across the country. The series launched on February 28 at St. Joseph’s Institute of Technology in Chennai, followed by events at institutions including Chitkara University in Punjab and Quantum University in Roorkee, creating a nationwide innovation trail. Through hands-on workshops, participants mastered MeTTa fundamentals and applied their learning to real-world AI applications across various fields, guided by expert mentors. Our partners at iCog Labs, represented by Samuel Wondimagegnehu and Yeabsira Nigusse, provided technical mentorship and support and also served as judges during the series. HackIndia 2025 represented an opportunity to extend our outreach to developer communities and showcase our decentralized AGI stack across diverse academic institutions. The initiative introduced MeTTa to over 150 universities, fostering new developer communities beyond traditional tech hubs and demonstrating that decentralized tools can empower AGI development across diverse regions.
Research paper? New market? Code snippet? Type what you need once → ASI:One finds the right agent to bring back exactly what you need. No tabs, no late-night scrolling. Our LLM will deliver everything for you with agentic intelligence. Try asi1.ai
⚡️ Panel Spotlight: AI Agents as On-Chain Decision-Makers Moderated by Adam Wozney @akashnet with speakers: • Janet Adams, COO @SingularityNET • Shaw Walters, Founder @ElizaLabs • Stefano Bury, Head of the US @virtuals_io • Yurii Kyparus, Co-founder @wallchain_xyz • Clara Tsao, Founding Officer @FilFoundation Join the discussion on how agents are reshaping governance, DeFi, and beyond. 📍Token2049 | Singapore| October 1
Join leading AI researchers, innovators, policymakers, and technologists, including @bengoertzel, @Plinz, and @romanyam, at the Beneficial AGI Summit 2025 to shape the future of AI and AGI. Register now to attend in Istanbul or watch online: bgisummit.io
Final call! 📣 Less than a week left to be part of what's #NextWithCUDOS. ⚡ Deploy inference instantly 📈 Earn compute credits + $FET rewards 🎟️ Early access 💸 $100K+ in rewards reserved for the first 1,000 verified users Read the full breakdown 👇 cudos.org/blog/last-chan…
In a conversation between CEO Dr. @bengoertzel and Edward @Snowden, questions arise about how governments approach AI regulation and the growing centralization of AI development. Snowden noted that corporations often invoke AI safety in ways that can slow smaller startups. He argued that current regulations favor larger organizations "by design," raising questions about how to ensure a fair and beneficial AI ecosystem. The implications extend beyond corporate boardrooms. Nations are approaching AI development as a strategic race, Snowden said, with early movers gaining a significant advantage in the years ahead. Advanced AI systems could reshape labor markets, military capabilities, and society at large. Snowden emphasized that the technology is evolving quickly and will have a profound impact. The Beneficial AGI Summit, taking place from October 21-23 in Istanbul, aims to address these challenges by bringing together diverse voices to steer AGI toward democratic and decentralized futures. Register now to secure your place and make your voice heard: luma.com/e4rx1p9b
We provide a high-level overview of Hyperon and PRIMUS, introducing researchers and developers to a complete seed-AGI capable of self-directed growth, immune to monopolistic control, and primed for cooperative co-evolution with humanity: link.springer.com/chapter/10.100…
We invite researchers and developers working toward AGI to join our upcoming MeTTa Coders session this Friday, September 5, at 3:30 PM UTC. This gathering brings together the Hyperon community to explore the latest developments in the MeTTa language for cognitive computations, collaborate, and advance practical use cases. To join the session and stay up to date with our AGI R&D initiatives, you can add our biweekly sessions to your calendar here: shorturl.at/R6cgQ
We invite researchers and developers working toward AGI to join our upcoming MeTTa Coders session this Friday, September 5, at 3:30 PM UTC. This gathering brings together the Hyperon community to explore the latest developments in the MeTTa language for cognitive computations, collaborate, and advance practical use cases. To join the session and stay up to date with our AGI R&D initiatives, you can add our biweekly sessions to your calendar here: shorturl.at/R6cgQ
