Nick Saraev and Jack Roberts discuss their informal tests of GLM 5.2, including a locally quantized modelModel quantization represents AI model values with fewer bits to reduce storage, memory use and often inference cost., coding outputs, speed and costs. They describe trade-offs and caution against relying on a single benchmarkEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria..
Nick Saraev and Jack Roberts compare model-routing approachesAI model routing sends each request to a model chosen for that request's complexity, cost, speed, privacy, or other requirements. and discuss how promptingPrompt engineering is the practice of designing and refining instructions, context, examples, and constraints to obtain useful AI outputs., review and user skill can affect results. Their assessments are personal observations rather than controlled benchmark results.
Nick Saraev and Jack Roberts discuss their views on AI-generated content, demonstrations of real work and audience relationships. Later audience questions touch on agent deployment and wearable interfaces.
Watch on YouTube
