I discuss my experience testing different AI systems prompting including Google Bard, OpenAI GPT-4 / GPT 3.5, Anthropic Claude 2, Llama 2, and Jasper to generate location-specific content. Most of this is based on the last 18 months of building out prompts, and now testing on models released over the last 4-6 weeks. Google Bard Released major update on July 13, 2023Prompt strategy: Long paragraphs, numbered tasks, multiple iterationsCouldn't produce high quality content without heavy editingIssues following instructions, needing reminders OpenAI GPT-4 Works well with conversational, transcribed promptAble to follow directions and produce high quality contentNo need for shot prompting OpenAI GPT-3.5 Uses revised GPT-4 prompt plus follow up to enforce formattingGets content production-ready after second promptQuality close to GPT-4 with additional data/content provided Anthropic Claude 2 No API access, using text interfaceRequired revising prompt structure significantlyXML tagging of data types improves contextBuilt-in prompt diagnosis/suggestions helpfulSingle prompt can produce high quality output Meta Llama 2 Free to use commercially if you have the hardwareExpected behavior similar to GPT-3.5GPT-4 prompt worked wellQuality closer to GPT-3.5 but better privacyCould refine with prompt chainingIssues following instructions precisely Jasper API Access useful for building AI toolsLong prompt length capabilityAppears to use GPT-4 or variantZero shot performs as well as GPT-4Able to produce high quality content easily Conclusion GPT-4 and Jasper produce quality results most easilyPleasantly surprised by Claude 2 quality and formatting of promptLlama 2 needs refinement to reach GPT-4 levelCurious about prompt strategies working across modelsFull show notes: https://opinionatedseo.com/2023/07/ai-prompting/