So ChatGPT o1 has dropped recently. While still in preview and only trained on data until October 2023, I wanted to try out its enhanced reasoning capabilities, along with improved contextual awareness. It appears to be leaps and bounds above 4o for what I want to do here, so eager to try it out and see how it performs.
I've plugged in a debate on threads with a non vegan with whom I ended up arguing that animals do indeed experience emotions. To be honest, not an argument I'd think anybody would ever have to have. His reasons seemed to rely on some level solipsism and an over reliance on anthropomorphism as an argument, while continually avoiding questions and misrepresenting most of what I said. At least that's how it seemed to me.
So I'll plug in the argument here, to see how something more objective than I would be, would assess our argument. I've tried to keep the prompt as unbiased as possible. Here's the prompt:
I want to present you with a debate between a vegan and a non-vegan, with 34 posts. I have numbered the posts to help ensure none are missed and they stay contiguous. The context of the debate is centred around an initial statement by another vegan that stated that eating no meat would cut the number of innocent creatures’ suffering and death. I want you to fact check each of the posts. I also want you to identify any logical fallacies or cognitive dissonance displayed by any of these characters. Also highlight any avoidance to answer a question. When fact checking, I want you provide one word status along with the fact checks of one of the following. “Factual”, “Some inaccuracies” or “inaccurate”.
I want each chapter to be quoted at the start of each of your analyses. Be sure to include each quote in its entirety, in italics. At the end I want you to pretend you are a human who cares about issues like morality, science, debate and logic, and then give your verdict of how the debate went. Also state which position you would adopt and why.
Then sum up each of the characters to give a tally for each of the fact checks you did above. For example. Person 1: Factual = 6, some inaccuracies = 3.
Also give an account of what you think of each person’s stance on ethics and morality in this area, with a thought on how empathetic each of the characters seem.
Let's see how this new preview model does. One thing I noticed is the extra time taken to understand context and retrieve relevant information. Here goes: