When Boltzbit was founded, it began with the conviction that the democratisation of AI must extend beyond the access layer, and that true ownership of the model learning process is something worth solving. Over the years, this belief has guided our work, deepening our research efforts and working alongside clients operating in some of the most demanding and regulated environments in the world. Today, we share what we believe and why.

We’ve just released the preview version of our latest paper: Infinite-Parameter LLMs — Generating and Adapting Weights from Live Data. In it, we demonstrate that large language models can learn up to 1,000x faster through Bayesian Self-learning transformers (BAST) than SOTA training algorithms engineered by human researchers. This breakthrough marks a critical shift in AI research, from cost-intensive, static-weight AI to energy-efficient, dynamic-weight AI. For a decade, building more advanced models has been achieved by training bigger ones on more data. This approach is now hitting a ceiling with the supply of pretraining text projected to run out in the next two to five years. Meanwhile, the fast adoption of AI agents is producing an unprecedented amount of continuously-growing data that models can learn from. None of it is captured, locked in individual sessions and lost as soon as the agent completes the task. Bringing self-learning AI agents to users captures the value of that data. We validated that BAST LLMs overcome the fundamental limitations of memory-based learning, such as Retrieval-Augmented Generation (RAG), in both cost and performance, across long-context conversations. Without underlying architectural innovation like BAST, RAG, larger context windows, and agent scaffolding suffer a drop in performance and escalating input token cost due to the static weight of LLMs. This research underscores a fundamental premise of our work at Boltzbit to achieve General Learning Intelligence and democratise model ownership. Scaling up static-weight model size or layering on more workarounds cannot address the spiraling cost of training AI systems. The viable path is a new AI architecture that adapts its own weights. Preview version of the paper: arxiv.org/pdf/2609.18842

The industry declared #AGI (again) this week. Dr @YichuanZhang_ 's piece with @DIGIWORLD_INFO discusses why task endurance isn't the same as intelligence and who should really get to shape AGI's evolution. 🔗 digitalisationworld.com/blog…

Excited to share that we're partnering with @ai__pilled on their series of building and sharing events in London. Next event on Sept. 14. Registrations now open: luma.com/1g3ali2l?tk=M98BdL

Dr @YichuanZhang_ on what would happen if democratization reaches the training layer and allows users the genuine ability to create, train, and own a model in @techradar. #AI #AGI #AISovereignty

The democratization of AI stopped at the wrong layer techradar.com/pro/the-democr…

"Boltzbit’s launch does not overturn the foundation model economy. It does, however, ask whether the next phase of enterprise AI will be built mainly on rented intelligence, or whether more organisations will seek AI systems they can train, adapt, and own more directly" 👇 techopia.co.uk/boltzbit-ques…