Matthew Berman compares the model’s reported evaluations with his own experience, describing extensive exploration, high token use and responses he found difficult to read. He discusses using less expensive models for simpler tasksModel selection chooses the AI model whose capability, quality, cost, speed, safety, and operating constraints best fit a task..
He describes frequent clarification requests and slow starts, then demonstrates browser simulations built with agent workflowsCode generation uses AI or another automated system to create source code from instructions, examples, schemas, or higher-level specifications.. The discussion treats the reported behavior as early personal testing rather than an independent benchmarkEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria..
Watch on YouTube
