3 ms·
Super interesting to see how - for the same set of prompts - GPT and Claude rely on just 50-60 domains while Perplexity and Gemini cite 100-140.
by maltelandwehr 1y ago
Super interesting to see how - for the same set of prompts - GPT and Claude rely on just 50-60 domains while Perplexity and Gemini cite 100-140.