Apart from playing around with these models to test their capabilities, how are people using them - either for personal use or in production settings?<p>EDIT: Specifically interested in how people are using the reasoning models (o1, DeepSeek) in ways that GPT-4 / 4o wasn't already good at.
Generic document summarization is my No. 1 use case.<p>Hopefully we should not see others here using them to "Stream-line" their "operations" but they will try anyway.