Boltzbit launches with new approach to AI built on live learning and model ownership
The London-based research company introduces General Learning Intelligence (GLI) as a new first-principles method to building models.
LONDON, July 30, 2026 – Boltzbit, the AI research company developing General Learning Intelligence (GLI), today unveiled its vision to make live-learning AI systems widely accessible. The company introduced General Learning Intelligence (GLI), the architecture behind this approach, alongside a new brand.
Recent advances in artificial intelligence have been driven by a small number of frontier AI labs building general-purpose, pre-trained models. These models are capital- and resource-intensive. Once deployed, they are static and unable to adapt to new conditions or improve through use without retraining.
Boltzbit’s live-learning technology addresses the fundamental challenge of AI training cost.
“It is undeniable that the current approach has enabled rapid progress and brought advanced AI capabilities into the mainstream,” said Dr. Yichuan Zhang, CEO and co-founder of Boltzbit. “But if AI is going to be genuinely democratised, access isn’t enough and the current trajectory is unsustainable. We need systems that perform and learn in the real world, and we need to democratise model ownership at user level. That’s what General Learning Intelligence is for.”
Boltzbit’s technology is already in production. It powers applications where live learning improves performance, reduces latency between insight and action, and enables more context-aware AI agents for complex decision-making.
“Training is the real bottleneck in AI, and we’ve seen what live-learning models are capable of in production through our work with select corporations,” said Dr. Jinli Hu, CTO and co-founder of Boltzbit. “AI adoption is moving quickly, but it is concentrating around pre-trained foundation models. That concentration has implications beyond competition. That’s why we are broadening access to our technology now, before it becomes too entrenched.”
The company’s brand launch marks the start of more active engagement with the market as Boltzbit works to widen access to its live-learning technology. Later this year, the firm plans to release its first product suite. It will bring together the components needed to create, train and host AI models, and build AI applications and agents on top of them.
ENDS
About Boltzbit
Boltzbit is an AI research company headquartered in London, developing General Learning Intelligence (GLI). Founded in 2020 by former Google AI researcher Dr Yichuan Zhang and former Microsoft AI researcher Dr Jinli Hu, the company builds on academic research into Boltzmann machines to develop generative models capable of live learning in production.
Its models are designed to address the limitations of traditional transformer pre-training or post-training by continuously adapting to new data and environments at inference time. Boltzbit’s technology is already in production, powering applications across financial services, data, and high-growth technology sectors.
The company is part of a new wave of research focused on moving beyond static, model-centric AI toward adaptive, context-driven intelligence, an approach referred to as AGI 2.0. For more information, visit www.boltzbit.com.
