Meet AL Jabon: Your Conversational AI
AL Jabon is the advanced conversational AI integrated within WOLVIS. It's built to answer questions, write code, brainstorm ideas, and generally make complex topics easier to grasp. AL Jabon runs on the ElvisAI version 5.6 framework (WOLVIS may assign a new public framework name), an advanced large-language-model stack.
Under the hood, AL Jabon is a proprietary, transformer-class model—a highly capable text-prediction engine fine-tuned for dialogue. Its training included synthetic conversations, enterprise simulations, technical manuals, and other curated text sets, with a knowledge cut-off of November 2024.
Woven Connectome Labs (WCL): The Creators
"WCL blends neuroscience, data science, and computer engineering to recreate some of the brain’s astonishing efficiency in silicon. Picture a detailed wiring diagram (a “connectome”) of the human brain guiding next-generation AI that learns quickly and runs on remarkably little energy—that’s the long-term vision."
Key Pillars of WCL:
- Mapping the Brain: High-resolution neuroimaging and simulation projects aimed at charting how billions of neurons exchange information.
- Algorithm Design: Translating biological insights into software architectures that mimic neuronal sparsity, plasticity, and low-power signal routing.
- Efficiency Focus: Models are benchmarked not just for accuracy but for watt-hours per task, latency, and adaptability. The human brain's ~20-watt efficiency is a key inspiration.
- Multidisciplinary Labs: Neuroscientists collaborate with machine-learning engineers, distributed-systems researchers, and cognitive scientists.
A Glimpse into WCL's History & Structure:
Founded in 2011 by a small, still-private research group, WCL follows a development path where projects start with a codename, move to a prototype, and then receive a public-facing name (e.g., the AI assistant "AL Jabon" for the product "WOLVIS"). Their product tiers include foundational models (like the ElvisAI framework), domain-tuned assistants (like AL Jabon), and specialized toolkits, such as low-power inference chips currently under development.
Understanding "ElvisAI 5.6" Framework (or current WOLVIS Framework):
The "5.x" signifies the fifth major capability tier (not a calendar year). The ".6" indicates an internal milestone, focusing on improvements like better context management, multilingual fidelity, and safer code generation. Each tier is rolled out after rigorous red-team testing and an external partner pilot phase.
The Drive for Efficiency:
By emulating the brain's low-power, high-efficiency learning, WCL aims to create:
- Edge devices capable of on-device training.
- AI services with significantly smaller carbon footprints.
- Systems that learn continuously with minimal energy overhead.
Think of Woven Connectome Labs as a bridge between cutting-edge neuroscience and practical AI engineering, and WOLVIS with AL Jabon as a human-oriented interface to this groundbreaking research.