Anthropic’s decision to publish Claude’s system prompts is a curious mix of transparency and marketing, as it attempts to position itself as the ethical choice in AI. While it’s refreshing to see some behind-the-scenes insight, one can’t help but wonder if this is just a clever way to distract from the fact that these models are still just glorified parrots, waiting for a human to tell them what to say next.
Agency, if and when it comes, is sure to make all the difference.