AI model cuts costs by 11 times
Pathway announced the launch of a new artificial intelligence model called "BDH-CQ," which is based on an unprecedented technology known as "Post-Transformers." Its goal is to enhance machines' ability to perform complex logical reasoning and significantly reduce operational costs. The new model represents rules and contextual patterns using digital vector technology, leading to lower computational consumption and improved operational efficiency. It can solve puzzles with an accuracy rate of nearly 30%, while reducing operating costs by 11 times compared to OpenAI's "GPT 5.6 Luna" model. The company's team plans to expand the model's architecture to include 600 billion parameters, aiming to apply it in cybersecurity and industrial operations. Reinforcing its effectiveness, independent experts emphasized that model engineering—not just size—is the key to developing AI's logical reasoning capabilities.
Notice: This Is an AI-Generated Summary
Comments (0)