How AI can Improve Prompt Creation

In the last three months alone, over ten papers outlining novel prompting techniques were published, boosting LLMs’ performance by a substantial margin. Two weeks ago, a groundbreaking paper from Microsoft demonstrated how a well-prompted GPT-4 outperforms Google’s Med-PaLM 2, a specialized medical model, solely through sophisticated prompting techniques. Yet, while our X and LinkedIn feeds buzz with ‘secret prompting tips’, a definitive, research-backed guide aggregating these advanced prompting strategies is hard to come by. This gap prevents LLM developers and everyday users from harnessing these novel frameworks to enhance performance and achieve more accurate results. https://lnkd.in/g7_6eP6y In this AI Tidbits Deep Dive, I outline six of the best and recent prompting methods: (1) EmotionPrompt - inspired by human psychology, this method utilizes emotional stimuli in prompts to gain performance enhancements (2) Optimization by PROmpting (OPRO) - a DeepMind innovation that refines prompts automatically, surpassing human-crafted ones. This paper discovered the “Take a deep breath” instruction that improved LLMs’ performance by 9%. (3) Chain-of-Verification (CoVe) - Meta's novel four-step prompting process that drastically reduces hallucinations and improves factual accuracy (4) System 2 Attention (S2A) - also from Meta, a prompting method that filters out irrelevant details prior to querying the LLM (5) Step-Back Prompting - encouraging LLMs to abstract queries for enhanced reasoning (6) Rephrase and Respond (RaR) - UCLA's method that lets LLMs rephrase queries for better comprehension and response accuracy Understanding the spectrum of available prompting strategies and how to apply them in your app can mean the difference between a production-ready app and a nascent project with untapped potential. Full blog post https://lnkd.in/g7_6eP6y

🧠 Designing AI That Thinks: Mastering Agentic Prompting for Smarter Results Have you ever used an LLM and felt it gave up too soon? Or worse, guessed its way through a task? Yeah, I've been there. Most of the time, the prompt is the problem. To get AI that acts more like a helpful agent and less like a chatbot on autopilot, you need to prompt it like one. Here are the three key components of an effective 🔁 Persistence: Ensure the model understands it's in a multi-turn interaction and shouldn't yield control prematurely. 🧾 Example: "You are an agent; please continue working until the user's query is resolved. Only terminate your turn when you are certain the problem is solved." 🧰 Tool Usage: Encourage the model to use available tools, especially when uncertain, instead of guessing. 🧾 Example:" If you're unsure about file content or codebase structure related to the user's request, use your tools to read files and gather the necessary information. Do not guess or fabricate answers." 🧠 Planning: Prompt it to plan before actions and reflect afterward. Prevent reactive tool calls with no strategy. 🧾 Example: "You must plan extensively before each function call and reflect on the outcomes of previous calls. Avoid completing the task solely through a sequence of function calls, as this can hinder insightful problem-solving." 💡 I've used this format in AI-powered research and decision-support tools and saw a clear boost in response quality and reliability. 👉 Takeaway: Agentic prompting turns a passive assistant into an active problem solver. The difference is in the details. Are you using these techniques in your prompts? I would love to hear what's working for you; leave a comment, or let's connect! #PromptEngineering #AgenticPrompting #LLM #AIWorkflow

Prompt optimization is becoming foundational for anyone building reliable AI agents Hardcoding prompts and hoping for the best doesn’t scale. To get consistent outputs from LLMs, prompts need to be tested, evaluated, and improved—just like any other component of your system This visual breakdown covers four practical techniques to help you do just that: 🔹 Few Shot Prompting Labeled examples embedded directly in the prompt help models generalize—especially for edge cases. It's a fast way to guide outputs without fine-tuning 🔹 Meta Prompting Prompt the model to improve or rewrite prompts. This self-reflective approach often leads to more robust instructions, especially in chained or agent-based setups 🔹 Gradient Prompt Optimization Embed prompt variants, calculate loss against expected responses, and backpropagate to refine the prompt. A data-driven way to optimize performance at scale 🔹 Prompt Optimization Libraries Tools like DSPy, AutoPrompt, PEFT, and PromptWizard automate parts of the loop—from bootstrapping to eval-based refinement Prompts should evolve alongside your agents. These techniques help you build feedback loops that scale, adapt, and close the gap between intention and output

Want to prompt like the top AI startups? 👇 YC shared tips how the top AI startups in their portfolio are prompting LLMs: Key learnings:  1/ Be Hyper-Specific & Detailed (The “Manager” Style) Treat your LLM like a new employee. Provide long, detailed prompts that define their role, task, constraints, and desired output. Example: Parahelp uses a 6+ page prompt for their AI customer support agent! 2/ Assign a Clear Role (Set a Persona) Start with: “You are a [role].” This sets the context, tone, and expected expertise. This helps the LLM adopt the desired style and reasoning for the tasks. 3/ Outline the Task + Provide the Steps Clearly state the LLM's primary task Break down complex tasks into a step-by-step plan. This Improves reliability and makes complex operations more manageable for the LLM. 4/ Structure Your Prompt (and Output) Use Markdown, bullet points, XML tags to structure your instructions Clear format helps with consistent and reliable outputs. Example: Parahelp, for instance, uses tags like <manager_verify> to enforce response format. 5/ Meta-Prompting (LLM, Improve Thyself) Yes, you can ask the LLM to help you write or refine prompts. Give it your current prompt. Ask it to make your prompt better or critique it. LLMs often suggest effective improvements you might not think of. 6/ Provide Examples For complex tasks, include a few high-quality examples of input-output pairs directly in the prompt. This improves the LLM's ability to understand and replicate desired behavior. Example: Jazzberry (AI bug finder) feeds hard examples to guide the LLM. 7/ Prompt Folding & Dynamic Generation Design prompts that generate specialized sub-prompts on the fly. Use this in multi-step workflows to break down complexity and adapt based on prior output. Example: A tool classification prompt that outputs a more targeted follow-up prompt like: “Now write a bug triage report for a frontend UI error.” 8/ Add an Escape Hatch Build fail-safes right into your prompt: Example: If you’re unsure or missing info, say ‘I don’t know’ and ask for clarification. This reduces hallucinations. Increases trust. 9/ Use Debug Info & Thinking Traces Ask the model to explain its reasoning “Include a section called ‘debug_info’ where you explain the logic behind your answer.” This is great for debugging and fine-tuning. 10/ Treat Evals Like Gold Yes, prompts matter. But evals are your most imp IP. Evals are essential for knowing why a prompt works and for iterating effectively. 11/ Consider Model "Personalities" & Distillation Different LLMs have different "personalities" Use the most powerful model to write and refine prompts. Then distill the optimized prompt for speed/post for production use. Know someone building AI agents?  Share this with them! Let’s level up our prompt engineering together 🔥 🔗 Source in comments #Startups #ArtificialIntelligence #PromptEngineering #AgenticAI #EnterpriseAI

Prompt engineering remains one the most effective alignment strategies because it allows developers to steer LLM behavior without modifying model weights, enabling fast, low-cost iteration. It also leverages the model’s pretrained knowledge and internal reasoning patterns, making alignment more controllable and interpretable through natural language instructions. However, it doesn’t come without cons, such as fragility of prompts (ex: changing one word can lead to different behavior), and scalability limits (ex: prompt engineer limits long chain reasoning capabilities). However, different tasks demand different prompting strategies, allowing you to select what best fit your business objectives, including budget constraints. If you're building with LLMs, you need to know when and how to use these. Let’s break them down: 1.🔸Chain of Thought (CoT) Teach the AI to solve problems step-by-step by breaking them into logical parts for better reasoning and clearer answers. 2.🔸ReAct (Reason + Act) Alternate between thinking and doing. The AI reasons, takes action, evaluates, and then adjusts based on real-time feedback. 3.🔸Tree of Thought (ToT) Explore multiple reasoning paths before selecting the best one. Helps when the task has more than one possible approach. 4.🔸Divide and Conquer (DnC) Split big problems into subtasks, handle them in parallel, and combine the results into a comprehensive final answer. 5.🔸Self-Consistency Prompting Ask the AI to respond multiple times, then choose the most consistent or commonly repeated answer for higher reliability. 6.🔸Role Prompting Assign the AI a specific persona like a lawyer or doctor to shape tone, knowledge, and context of its replies. 7.🔸Few-Shot Prompting Provide a few good examples and the AI will pick up the pattern. Best for structured tasks or behavior cloning. 8.🔸Zero-Shot Chain of Thought Prompt the AI to “think step-by-step” without giving any examples. Great for on-the-fly reasoning tasks. Was this type of guide useful to you? Let me know below. Follow for plug-and-play visuals, cheat sheets, and step-by-step agent-building guides. #genai #promptengineering #artificialintelligence

Which is it: use LLMs to improve the prompt, or is that over-engineering? By now, we've all seen a 1000 conflicting prompt guides. So, I wanted to get back to the research: • What do actual studies say? • What actually works in 2025 vs 2024? • What do experts at OpenAI, Anthropic, & Google  say? I spent the past month in Google Scholar, figuring it out. I firmed up the learnings with Miqdad Jaffer at OpenAI. And I'm ready to present: "The Ultimate Guide to Prompt Engineering in 2025: The Latest Best Practices." https://lnkd.in/d_qYCBT7 We cover: 1. Do You Really Need Prompt Engineering? 2. The Hidden Economics of Prompt Engineering 3. What the Research Says About Good Prompts 4. The 6-Layer Bottom-Line Framework 5. Step-by-step: Improving Your Prompts as a PM 6. The 301 Advanced Techniques Nobody Talks About 7. The Ultimate Prompt Template 2.0 8. The 3 Most Common Mistakes Some of my favorite takeaways from the research: 1. It's not just revenue, but cost You have to realize that APIs charge by number of input and output tokens. An engineered prompt can deliver the same quality with 76% cost reduction. We're talking $3,000 daily vs $706 daily for 100k calls. 2. Chain-of-Table beats everything else This new technique gets 8.69% improvement on structured data by manipulating table structure step-by-step instead of reasoning about tables in text. For things like financial dashboards and data analysis tools, it's the best. 3. Few-shot prompting hurts advanced models OpenAI's o1 and DeepSeek's R1 actually perform worse with examples. These reasoning models don't need your sample outputs - they're smart enough to figure it out themselves. 4. XML tags boost Claude performance Anthropic specifically trained Claude to recognize XML structure. You get 15-20% better performance just by changing your formatting from plain text to XML tags. 5. Automated prompt engineering destroys manual AI systems create better prompts in 10 minutes than human experts do after 20 hours of careful optimization work. The machines are better at optimizing themselves than we are. 6. Most prompting advice is complete bullshit Researchers analyzed 1,500+ academic papers and found massive gaps between what people claim works and what's actually been tested scientifically. And what about Ian Nuttal's tweet? Well, Ian's right about over-engineering. But for products, prompt engineering IS the product. Bolt hit $50M ARR via systematic prompt engineering. The Key? Knowing when to engineer vs keep it simple.

Some of the best AI breakthroughs we’ve seen came from small, focused teams working hands-on, with structured inputs and the right prompting. Here’s how we help clients unlock AI value in days, not months: 1. Start with a small, cross-functional team (4–8 people) 1–2 subject matter experts (e.g., supply chain, claims, marketing ops) 1–2 technical leads (e.g., SWE, data scientist, architect) 1 facilitator to guide, capture, and translate ideas Optional: an AI strategist or business sponsor 2. Context before prompting - Capture SME and tech lead deep dives (recorded and transcribed) - Pull in recent internal reports, KPIs, dashboards, and documentation - Enrich with external context using Deep Research tools: Use OpenAI’s Deep Research (ChatGPT Pro) to scan for relevant AI use cases, competitor moves, innovation trends, and regulatory updates. Summarize into structured bullets that can prime your AI. This is context engineering: assembling high-signal input before prompting. 3. Prompt strategically, not just creatively Prompts that work well in this format: - “Based on this context [paste or refer to doc], generate 100 AI use cases tailored to [company/industry/problem].” - “Score each idea by ROI, implementation time, required team size, and impact breadth.” - “Cluster the ideas into strategic themes (e.g., cost savings, customer experience, risk reduction).” - “Give a 5-step execution plan for the top 5. What’s missing from these plans?” - “Now 10x the ambition: what would a moonshot version of each idea look like?” Bonus tip: Prompt like a strategist (not just a user) Start with a scrappy idea, then ask AI to structure it: - “Rewrite the following as a detailed, high-quality prompt with role, inputs, structure, and output format... I want ideas to improve our supplier onboarding process with AI. Prioritize fast wins.” AI returns something like: “You are an enterprise AI strategist. Based on our internal context [insert], generate 50 AI-driven improvements for supplier onboarding. Prioritize for speed to deploy, measurable ROI, and ease of integration. Present as a ranked table with 3-line summaries, scoring by [criteria].” Now tune that prompt; add industry nuances, internal systems, customer data, or constraints. 4. Real examples we’ve seen work: - Logistics: AI predicts port congestion and auto-adjusts shipping routes - Retail: Forecasting model helps merchandisers optimize promo mix by store cluster 5. Use tools built for context-aware prompting - Use Custom GPTs or Claude’s file-upload capability - Store transcripts and research in Notion, Airtable, or similar - Build lightweight RAG pipelines (if technical support is available) - Small teams. Deep context. Structured prompting. Fast outcomes. This layered technique has been tested by some of the best in the field, including a few sharp voices worth following, including Allie K. Miller!

Prompt Smarter, Research Better. I summarized what 6 weeks of Prompt engineering class taught me (save this guide for free) Ai Prompting is a research skill. Just like coding, writing, or presenting, mastering how to prompt AI will soon be a core academic asset. Remember, Quality prompt = Quality output. Garbage in = Garbage out. — ➊ 𝗔𝗹𝘄𝗮𝘆𝘀 𝗦𝗲𝘁 𝗧𝗵𝗲 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 LLMs respond better with clarity. — ➋ 𝗨𝘀𝗲 𝗥𝗼𝗹𝗲-𝗕𝗮𝘀𝗲𝗱 𝗣𝗿𝗼𝗺𝗽𝘁𝘀 Frame the AI as someone you want it to emulate. ➤ “You are a research mentor with expertise in social epidemiology. — ➌ 𝗔𝘀𝗸 𝗪𝗲𝗹𝗹-𝗦𝗰𝗼𝗽𝗲𝗱 𝗤𝘂𝗲𝘀𝘁𝗶𝗼𝗻𝘀 ↳ Break down complex queries into smaller parts. — ➍ 𝗟𝗲𝘃𝗲𝗿𝗮𝗴𝗲 𝗣𝗿𝗼𝗺𝗽𝘁 𝗖𝗵𝗮𝗶𝗻𝗶𝗻𝗴 (Step-by-step logic) ➤ Layer your prompts. ➎ 𝗘𝗱𝘂𝗰𝗮𝘁𝗲 𝗧𝗵𝗲 𝗠𝗼𝗱𝗲𝗹 (if needed) ➤ LLMs are generalists. ➤ Feed them your working definitions. ➤ “For this study, we define ‘access to care’ as…” — ➏ 𝗥𝗲𝗾𝘂𝗲𝘀𝘁 𝗢𝘂𝘁𝗽𝘂𝘁 𝗙𝗼𝗿𝗺𝗮𝘁𝘀 Be specific about what format you want: → “Summarize in bullet points.” → “Give examples in APA format.” — ➐ 𝗧𝗲𝘀𝘁 𝗳𝗼𝗿 𝗕𝗶𝗮𝘀, 𝗙𝗮𝗰𝘁𝘂𝗮𝗹 𝗘𝗿𝗿𝗼𝗿𝘀 & 𝗛𝗮𝗹𝗹𝘂𝗰𝗶𝗻𝗮𝘁𝗶𝗼𝗻 Always verify AI-generated facts. ➤ Cross-check references. ➤ Ask follow-up: “What is the source of that claim?” ———————— Pro-Tip: Save Your Prompt Templates Create prompt banks for recurring tasks: ♻️Repost for others…. #AI #research #academia

Ready to 10x your AI prompting? Here’s the 5-part framework I use when I absolutely need a high-quality answer from AI: ✅ 1. Start with who you are I’ll say: “I’m a VC based in NY and Boston. I focus on seed-stage AI companies. I’ve been doing this for 25+ years.” Your perspective matters. The model needs a lens to think through. ✅ 2. Explain what you’re solving — and why Don’t just ask a question. Share the stakes: What’s the decision? Why does it matter? Who is it for? ✅ 3. Feed it real signal Transcripts. Research. Emails. Drafts. Anything relevant. The more information you give, the smarter the response becomes. ✅ 4. Be exact about the output Tell it what you want back: “3-paragraph summary. Include a pros/cons list. Write with the level of sophistication of the smartest investor in the world.” AI doesn’t guess well. You have to show it the target. ✅ 5. Prompt it to be a little mean The models are way too nice. I will instruc it: “Be critical. Don’t sugarcoat it. Hold me to the highest standard.” You’ll be shocked how much better the feedback gets. Most people give up on AI after a weak answer. But great prompting takes work. It’s a skill — and it’s quickly becoming a competitive edge. AI is not going to replace you, but someone who masters AI prompting WILL.

Prompting tips from someone that spends probably $13k+ per month on OpenAI API calls. I'll break the tips into chatGPT user interface tips as well as API tips. My bias is of course going to be about outbound sales and cold email because this is where we spend from and 100% of this spend is on 4o mini API calls. Chat GPT Prompting Tips 1. Use transcription as much as possible. Straight in the UI or use whisprflow(dot)ai (can't tag them for some reason). I personally get frustrated with a prompt when I'm typing it out vs. talking and can add so much more detail. 2. Got this one from Yash Tekriwal 🤔 - When you're working on something complex like a deep research request or something you want o3 to run or analyzing a lot of data, ask chatgpt to give you any follow up questions it might have before it runs fully. Helps you increase your prompt accuracy like crazy. 3. I've found that o3 is pretty good at building simple automations in make as well so we will ask it to output what we want in a format that we can input into make and often we can build automations just by explaining what we need and then plugging in our logins in Make. API prompting tips 1. Throwing back to the Chat GPT UI, but we will often create our complex prompts in the user interface first and then bring it into the API via Clay asking ChatGPT along the way on how to improve the prompt and help us think of edge cases. This can take any team member to a prompting pro immediately. 2. Examples are your best friend. Giving examples of what you would want the output to be is how we can get our outputs to be the same format and not put "synergies" in every email we are sending. I tell the team, minimum 2 examples for single line outputs. 4 examples for anything more complex than that. 6 examples for industry tagging because that gets so odd. Save on costs by putting some real examples in your system prompt. 3. Request the output in JSON. It keeps everything more uniform in the format you need. 4. Speaking of JSON, ask the API to prove to you why it thinks what it thinks and then output the answer. Especially for company category tagging, I find this works really well. I see this greatly increase the accuracy of our results for 2 reasons. I think if AI has to take the extra second to prove to you why a company is an ecommerce brand, the results are demonstrably better. This is just a guess, but I also think that because LLMs basically work by guessing what the next best word is, if you have it tell you why it thinks something is a certain industry and then it gives the output, I think it's much more likely to be correct. Anything else you've found?