I'm trying to be constructive in a destructive manner here, after reading the latest BRN newsletter. Looking at the letter, and links from it, from the point of view of a potential investor, makes me think such an investor would find BRN is too hard to understand, and so would not invest. That's a factor suppressing the share price.
Everywhere you read, it is certainly concise to the initiated, but written full of tech terms.
ChatGPT to the rescue! I asked it (by copying and pasting from the newsletter)
"What does the following mean "ENNs addresses the significant energy and computational demands of current AI systems. This technology excels in handling sequential and continuous data streams for edge and event-based processing, making it perfect for smart home, automotive, healthcare, and industrial applications.TENNs integrates state space models developed in parallel and generalizes convolution kernels to handle various transformer tasks, such as processing language, time-series, and spatiotemporal data. This broadens its applicability while significantly reducing energy consumption compared to traditional models."
It answeredENNs (Edge Neural Networks):
Energy and Computational Efficiency:ENNs are designed to significantly reduce the energy and computational resources required by current AI systems. This is important because traditional AI models can be very resource-intensive, leading to high operational costs and potential environmental impact.Handling Sequential and Continuous Data Streams:ENNs are particularly good at processing data that comes in sequences or continuously. This is crucial for applications where data is not static but changes over time, such as monitoring systems or real-time analytics.Edge and Event-Based Processing:Edge processing refers to performing data analysis and computation close to the source of data (e.g., on devices like sensors or smartphones) rather than sending it to a centralized server. This reduces latency and bandwidth usage.Event-based processing involves reacting to specific events or changes in the environment, which is efficient for real-time applications.Application Areas:ENNs are ideal for use in smart home technologies (like smart thermostats and security systems), automotive (such as autonomous driving and vehicle monitoring), healthcare (patient monitoring and diagnostics), and industrial settings (predictive maintenance and process optimization).TENNs (Transformer Enhanced Neural Networks):
State Space Models and Convolution Kernels:TENNs integrate state space models, which are mathematical models used to describe systems that change over time. These models help TENNs to effectively manage complex and dynamic data.The technology generalizes convolution kernels (a core component of convolutional neural networks, typically used for image processing) to be applicable to various tasks handled by transformers. Transformers are a type of model architecture used prominently in natural language processing (NLP) and other tasks requiring sequence data handling.Transformer Tasks:TENNs can process different types of data including language (NLP tasks like translation and sentiment analysis), time-series data (financial forecasting, sensor data analysis), and spatiotemporal data (data that includes both spatial and temporal dimensions, like video processing or geographic information systems).Broader Applicability and Energy Efficiency:By integrating these advanced techniques, TENNs can be used across a wider range of applications compared to traditional models. This flexibility makes them suitable for many AI tasks.Importantly, TENNs achieve this broader applicability while significantly reducing energy consumption, which is a major advantage in making AI more sustainable and cost-effective.In summary, both ENNs and TENNs are advanced AI technologies that address the high energy and computational demands of traditional AI systems. They are particularly suited for applications requiring real-time, continuous data processing, and they offer significant improvements in energy efficiency and versatility across various domains.
(20min delay)