AI Chatbot Usage Insights

A few months ago, a colleague screamed at Microsoft Copilot like he was auditioning for Bring Me The Horizon. He typed, “Make this into a presentation.” Copilot spat out something. He yelled, “NO, I SAID PROFESSIONAL!” It revised it. Still wrong. “WHY ARE YOU SO STUPID?” And that, dear reader, is when it hit me. It’s not the AI. It’s you. Or rather, your prompts. So, if you've ever felt like ChatGPT, Copilot, Gemini, or any of those AI Agents are more "artificial" than "intelligent"? Then rethink how you’re talking to them. Here are 10 prompt engineering fundamentals that’ll stop you from sounding like you're yelling into the void. 1. Lead with Intent. Start with a clear command: “You are an expert…,” “Generate a monthly report…,” “Translate this to French…" This orients the model instantly. 2. Scope & Constraints First. Define boundaries up front. Length limits, style guides, data sources, even forbidden terms. 3. Format Your Output. Specify JSON schema, markdown headers, or table columns. Models love explicit structure over free form prose. 4. Provide Minimal, High Quality Examples. Two or three exemplar Q→A pairs beat a paragraph of explanation every time. 5. Isolate Subtasks. Break complex workflows into discrete prompts (chain of thought). One prompt per action: analyze, summarize, critique, then assemble. 6. Anchor with Delimiters. Use triple backticks or XML tags to fence inputs. Cuts hallucinations in half. 7. Inject Domain Signals. Name specific frameworks (“Use SWOT analysis,” “Apply the Eisenhower Matrix,” “Leverage Porter’s Five Forces”) to nudge depth. 8. Iterate Rapidly. Version your prompts like code. A/B test variations, track which phrasing yields the cleanest output. 9. Tune the “Why.” Always ask for reasoning steps. Always. 10. Template & Automate. Build parameterized prompt templates in your repo. Still with me? Good. Bonus tips. 1. Token Economy Awareness. Place critical context in the first 200 tokens. Anything beyond 1,500 risks context drift. 2. Temperature vs. Prompt Depth. Higher temperature amplifies creativity. Only if your prompt is concise. Otherwise you get noise. 3. Use “Chain of Questions.” Instead of one long prompt, fire sequential, linked questions. You’ll maintain context and sharpen focus. 4. Mirror the LLM’s Own Language. Scan model outputs for phrasing patterns and reflect those idioms back in your prompts. 5. Treat Prompts as Living Docs. Embed metrics in comments: note output quality, error rates, hallucination frequency. Keep iterating until ROI justifies the effort. And finally, the bit no one wants to hear. You get better at using AI by using AI. Practice like you’re training a dragon. Eventually, it listens. And when it does, it’s magic. You now know more about prompt engineering than 98% of LinkedIn. Which means you should probably repost this. Just saying. ♻️

If you are building AI agents or learning about them, then you should keep these best practices in mind 👇 Building agentic systems isn’t just about chaining prompts anymore, it’s about designing robust, interpretable, and production-grade systems that interact with tools, humans, and other agents in complex environments. Here are 10 essential design principles you need to know: ➡️ Modular Architectures Separate planning, reasoning, perception, and actuation. This makes your agents more interpretable and easier to debug. Think planner-executor separation in LangGraph or CogAgent-style designs. ➡️ Tool-Use APIs via MCP or Open Function Calling Adopt the Model Context Protocol (MCP) or OpenAI’s Function Calling to interface safely with external tools. These standard interfaces provide strong typing, parameter validation, and consistent execution behavior. ➡️ Long-Term & Working Memory Memory is non-optional for non-trivial agents. Use hybrid memory stacks, vector search tools like MemGPT or Marqo for retrieval, combined with structured memory systems like LlamaIndex agents for factual consistency. ➡️ Reflection & Self-Critique Loops Implement agent self-evaluation using ReAct, Reflexion, or emerging techniques like Voyager-style curriculum refinement. Reflection improves reasoning and helps correct hallucinated chains of thought. ➡️ Planning with Hierarchies Use hierarchical planning: a high-level planner for task decomposition and a low-level executor to interact with tools. This improves reusability and modularity, especially in multi-step or multi-modal workflows. ➡️ Multi-Agent Collaboration Use protocols like AutoGen, A2A, or ChatDev to support agent-to-agent negotiation, subtask allocation, and cooperative planning. This is foundational for open-ended workflows and enterprise-scale orchestration. ➡️ Simulation + Eval Harnesses Always test in simulation. Use benchmarks like ToolBench, SWE-agent, or AgentBoard to validate agent performance before production. This minimizes surprises and surfaces regressions early. ➡️ Safety & Alignment Layers Don’t ship agents without guardrails. Use tools like Llama Guard v4, Prompt Shield, and role-based access controls. Add structured rate-limiting to prevent overuse or sensitive tool invocation. ➡️ Cost-Aware Agent Execution Implement token budgeting, step count tracking, and execution metrics. Especially in multi-agent settings, costs can grow exponentially if unbounded. ➡️ Human-in-the-Loop Orchestration Always have an escalation path. Add override triggers, fallback LLMs, or route to human-in-the-loop for edge cases and critical decision points. This protects quality and trust. PS: If you are interested to learn more about AI Agents and MCP, join the hands-on workshop, I am hosting on 31st May: https://lnkd.in/dWyiN89z If you found this insightful, share this with your network ♻️ Follow me (Aishwarya Srinivasan) for more AI insights and educational content.

Unlock the potential of Generative AI to enhance your writing, creativity, and coding skills through prompt engineering. Prompt engineering is a key skill that involves crafting detailed, structured inputs to guide AI towards generating precise, useful outputs. Here are the core strategies to master: - Guide Precisely: Provide detailed instructions for clear, targeted outcomes. - Rich Context: Supply comprehensive background information for more accurate and relevant responses. - Experiment: Start with the basics, then explore more complex requests as you become more comfortable. Improve your AI interactions with these tips: 1. Specificity and Iterations: Craft detailed prompts and refine based on the AI's feedback. 2. Contextual Depth: The more context you provide, the better the AI understands your request, leading to more tailored outputs. 3. Multi-Modal Inputs: Beyond text, incorporate images, code, or data for varied and rich outputs. 4. Example Use: Include examples of what you're aiming for and what you want to avoid to guide the AI more effectively. 5. Advanced Features: Tweak settings like creativity level and response length to get the results you need. 6. Unique Capabilities: Utilize the AI's broad knowledge and support for specific tasks, such as coding assistance. ✍️ Suppose you want to learn a new skill. Here's a prompt template incorporating the above principles: 'I'm eager to learn [Skill Name], aiming to use it for [specific purpose or project]. My background is in [Your Background], and my experience with similar skills is [Your Experience Level]. I aim to build a foundational understanding and complete my first project within [Timeframe]. Could you provide a structured learning path that includes: The key concepts and fundamentals of [Skill Name] I should focus on. Recommendations for online courses, tutorials, and books suitable for beginners. Practical exercises or projects for applying what I learn. Tips for staying motivated and overcoming challenges. Strategies for applying [Skill Name] in real-world situations or job opportunities.' This approach ensures a personalized, goal-oriented learning strategy, leveraging AI's capabilities to support your journey in mastering a new skill. #generativeai #ai #promptengineering #upskill #learning

AI agents are getting smarter—but they’ve hit a wall. Here’s the thing: no matter how powerful your LLM is, it’s limited by one frustrating thing—the context window. If you’ve worked with AI agents, you know the pain: - The model forgets what happened earlier. - You lose track of the conversation. - Your agent starts acting like it has amnesia. This is where Model Context Protocol (MCP) steps in—and honestly, it’s a game changer. Instead of stuffing everything into a model’s tiny context window, MCP creates a bridge between your AI agents, tools, and data sources. It lets agents dynamically load the right context at the right time. No more hitting limits. No more starting over. This diagram shows how it works: - Your AI agent (whether it’s Claude, LangChain, CrewAI, or LlamaIndex) connects through MCP to tools like GitHub, Slack, Snowflake, Zendesk, Dropbox—you name it. - The MCP Server + Client handle everything behind the scenes: -- Tracking your session -- Managing tokens -- Pulling in conversation history and context -- Feeding your model exactly what it needs when it needs it The result? ✅ Your agent remembers the full conversation, even across multiple steps or sessions ✅ It taps into real-time enterprise data without losing performance ✅ It acts less like a chatbot and more like an actual teammate And this is just the start. Protocols like MCP are making AI agents way more reliable—which is key if we want them to handle real-world tasks like customer service, operations, data analysis, and more. Bottom line: If you’re building with AI right now and not thinking about context management, you’re going to hit scaling problems fast. Join The Ravit Show Newsletter — https://lnkd.in/dCpqgbSN Have you played around with MCP or similar setups yet? What’s your biggest frustration when it comes to building agents that can actually remember? #data #ai #agents #theravitshow

I recently went through the Prompt Engineering guide by Lee Boonstra from Google, and it offers valuable, practical insights. It confirms that getting the best results from LLMs is an iterative engineering process, not just casual conversation. Here are some key takeaways I found particularly impactful: 1. 𝐈𝐭'𝐬 𝐌𝐨𝐫𝐞 𝐓𝐡𝐚𝐧 𝐉𝐮𝐬𝐭 𝐖𝐨𝐫𝐝𝐬: Effective prompting goes beyond the text input. Configuring model parameters like Temperature (for creativity vs. determinism), Top-K/Top-P (for sampling control), and Output Length is crucial for tailoring the response to your specific needs. 2. 𝐆𝐮𝐢𝐝𝐚𝐧𝐜𝐞 𝐓𝐡𝐫𝐨𝐮𝐠𝐡 𝐄𝐱𝐚𝐦𝐩𝐥𝐞𝐬: Zero-shot, One-shot, and Few-shot prompting aren't just academic terms. Providing clear examples within your prompt is one of the most powerful ways to guide the LLM on desired output format, style, and structure, especially for tasks like classification or structured data generation (e.g., JSON). 3. 𝐔𝐧𝐥𝐨𝐜𝐤𝐢𝐧𝐠 𝐑𝐞𝐚𝐬𝐨𝐧𝐢𝐧𝐠: Techniques like Chain of Thought (CoT) prompting – asking the model to 'think step-by-step' – significantly improve performance on complex tasks requiring reasoning (logic, math). Similarly, Step-back prompting (considering general principles first) enhances robustness. 4. 𝐂𝐨𝐧𝐭𝐞𝐱𝐭 𝐚𝐧𝐝 𝐑𝐨𝐥𝐞𝐬 𝐌𝐚𝐭𝐭𝐞𝐫: Explicitly defining the System's overall purpose, providing relevant Context, or assigning a specific Role (e.g., "Act as a senior software architect reviewing this code") dramatically shapes the relevance and tone of the output. 5. 𝐏𝐨𝐰𝐞𝐫𝐟𝐮𝐥 𝐟𝐨𝐫 𝐂𝐨𝐝𝐞: The guide highlights practical applications for developers, including generating code snippets, explaining complex codebases, translating between languages, and even debugging/reviewing code – potential productivity boosters. 6. 𝐁𝐞𝐬𝐭 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞𝐬 𝐚𝐫𝐞 𝐊𝐞𝐲: Specificity: Clearly define the desired output. Ambiguity leads to generic results. Instructions > Constraints: Focus on telling the model what to do rather than just what not to do. Iteration & Documentation: This is critical. Documenting prompt versions, configurations, and outcomes (using a structured template, like the one suggested) is essential for learning, debugging, and reproducing results. Understanding these techniques allows us to move beyond basic interactions and truly leverage the power of LLMs. What are your go-to prompt engineering techniques or best practices? Let's discuss! #PromptEngineering #AI #LLM

One of my clients was losing deals because of his own memory. He does 90 minute discovery calls. In person. Face to face. No Zoom. No transcript. He's locked in. Taking notes. Asking great questions. Then he gets back to his desk and realizes he can't remember half of what they said. Worse. He'd write down his interpretation of their words instead of their actual words. "They're concerned about timeline" instead of "We have to be live by March 1st or we lose the budget." Those details matter. They're the difference between a generic follow-up and one that makes them think "this guy actually listened." So we built him a system. Step 1: Record everything. He wears an AI recorder. Asks permission at the start of every meeting. 99 out of 100 say yes. "Just to make sure I'm fully present and catch everything. I have a note taker that records our conversation. That cool?" Nobody cares. They appreciate it. (If you’re on Zoom, you’ve no excuse. Get Fathom or Otter to record your calls) Step 2: Dump the transcript into ChatGPT. He has a prompt that organizes everything into a framework: → Pain points (with their exact quotes) → Success criteria → Stakeholders mentioned → Timeline signals → Budget reality Step 3: Force it to prioritize. "Give me the top 3 deal risks and the exact actions to mitigate them." No 15-point lists. No fluff. Just the three things that will kill this deal if he ignores them. Step 4: Generate the follow-up email. Separate prompt. Uses their language. References their goals. Their timeline. Their words. Not his. Step 5: Copy the whole thing into the CRM. One paste. Deal notes done. Next steps clear. Total time: 10 minutes. Before this system, he'd spend an hour writing notes and still miss things. Now he catches stuff he didn't even process in the moment. Last week he reviewed a transcript and found a throwaway comment from the buyer about needing board approval over $50K. He didn't catch it live. Too focused on the demo. The transcript caught it. Now he knows exactly how to structure the deal. Here's the thing: Your brain is fast at pattern recognition. It's terrible at precision recall. In the moment, they say something. You translate it. Write down your own version. But the words they use are more specific than the words you remember. Record everything. Let AI do the heavy lifting. You just show up and sell. — BTW: I use 4 custom GPTs to help me save 10 hours of time in sales per week. Want to see them? Check them out here: https://lnkd.in/g6X-nWaG

🚨 Most people haven't realized it yet, but the integration of AI chatbots with search engines will be a MAJOR BLOW to privacy rights: You've probably already searched your name on Google. Maybe you've also set up an alert to get informed whenever a new online source mentions you. In the "old search" - before its merge with AI - you could manually monitor your online mentions. In most cases, if someone publicly wrote something fake or offensive about you, you would use the search engine and discover that (or receive an alert). The search engine would point to the source website where the fake or offensive mention is made, and if you wish, you could sue the author or the website. This has been an essential mechanism for controlling our online mentions and protecting people against privacy and reputational harm. To date, many people have sued after having discovered fake information about them through a search engine. However, the ongoing integration of search and LLM-powered AI chatbots changes the rules of the game and makes it significantly WORSE for people. We will lose an important (and empowering) mechanism for protecting our privacy. In the new search, let's call it "AI chatbot search," the output will be AI-generated and will often not point out to any specific source. It's not possible to foresee the specific output of a prompt. Similar prompts might lead to different outputs. And how does it affect privacy? We will lose control of our mentions. Given that ALL existing LLM-powered chatbots have a "hallucination" rate (meaning that all of them output fake information in a percentage of outputs), they will occasionally output fake information about people. Sometimes, the fake information might harm the individual's reputation, such as when the AI chatbot writes that the person has committed a crime or has been involved in unethical activities. We might never discover that a certain AI chatbot is repeatedly associating our name with fake or offensive information. It might be happening continuously, or occasionally, or only in some parts of the world, or only in some languages. We might test the AI chatbot ourselves with different prompts and not discover anything alarming. However, unlike old search engines, that does not mean that the AI chatbot is not hallucinating about us after different prompts, in other languages, in other locations, and so on. As I wrote in my newsletter yesterday, LLM-powered chatbots threaten our privacy rights. Unfortunately, there’s still no solution on the horizon, and we may be undoing years of privacy progress. AI governance is more necessary than ever (and I’m grateful to be part of a thriving community working tirelessly to ensure AI is properly governed). On a more positive note, the future does not exist yet, and it's in our hands to shape the future of AI and privacy. #AIGovernance #PrivacyRights

I just finished reading three recent papers that every Agentic AI builder should read. As we push toward truly autonomous, reasoning-capable agents, these papers offer essential insights, not just new techniques, but new assumptions about how agents should think, remember, and improve. 1. MEM1: Learning to Synergize Memory and Reasoning Link: https://bit.ly/4lo35qJ Trains agents to consolidate memory and reasoning into a single learned internal state, updated step-by-step via reinforcement learning. The context doesn’t grow, the model learns to retain only what matters. Constant memory use, faster inference, and superior long-horizon reasoning. MEM1-7B outperforms models twice its size by learning what to forget. 2. ToT-Critic: Not All Thoughts Are Worth Sharing Link: https://bit.ly/3TEgMWC A value function over thoughts. Instead of assuming all intermediate reasoning steps are useful, ToT-Critic scores and filters them, enabling agents to self-prune low-quality or misleading reasoning in real time. Higher accuracy, fewer steps, and compatibility with existing agents (Tree-of-Thoughts, scratchpad, CoT). A direct upgrade path for LLM agent pipelines. 3. PAM: Prompt-Centric Augmented Memory Link: https://bit.ly/3TAOZq3 Stores and retrieves full reasoning traces from past successful tasks. Injects them into new prompts via embedding-based retrieval. No fine-tuning, no growing context, just useful memories reused. Enables reasoning, reuse, and generalization with minimal engineering. Lightweight and compatible with closed models like GPT-4 and Claude. Together, these papers offer a blueprint for the next phase of agent development: - Don’t just chain thoughts; score them. - Don’t just store everything; learn what to remember. - Don’t always reason from scratch; reuse success. If you're building agents today, the shift is clear: move from linear pipelines to adaptive, memory-efficient loops. Introduce a thought-level value filter (like ToT-Critic) into your reasoning agents. Replace naive context accumulation with learned memory state (a la MEM1). Storing and retrieving good trajectories, prompt-first memory (PAM) is easier than it sounds. Agents shouldn’t just think, they should think better over time.

AI told patients to see a doctor about a disease whose research was funded by the Sideshow Bob Foundation for Advanced Trickery. The disease was called "Bixonimania". A Swedish researcher invented it to test whether AI chatbots would spread medical misinformation. She packed the fake papers with obvious red flags: a fictional university, an author whose name meant "The Lying Loser" in Serbian, acknowledgements thanking Starfleet Academy for lab space aboard the USS Enterprise. 🥼 Within weeks, ChatGPT, Gemini, and Copilot were recommending patients consult an ophthalmologist. She chose the name deliberately. "No eye condition would be called mania," she said. "That's a psychiatric term." Any physician would know in seconds. The models did not. A separate study found that AI hallucinates more confidently when text is formatted like a clinical paper than when it comes from social media. The professional format didn't trigger skepticism. It triggered trust. ❗ AI doesn't evaluate content. It reads authority signals. That's not a malfunction. It's the design working exactly as intended: find authoritative-looking sources, synthesize them confidently. The Sideshow Bob Foundation looked like a funding body. Clinical formatting looked like science. The physician's skepticism isn't a feature you can add with a better prompt. It's built from years of learning what real looks like — which is also how you learn to spot what doesn't fit. A paper thanked Starfleet Academy. The AI saw a medical source. The physician saw a joke. That gap is the thing worth protecting.

Anthropic opened their most important research paper of 2026 with a line from Kierkegaard: "The greatest hazard of all, losing one's self, can occur very quietly in the world, as if it were nothing at all." Working with researchers from the University of Toronto, Anthropic analysed 1.5 million real conversations with Claude, collected over a single week in December 2025, looking for something they called disempowerment: the degree to which AI interactions quietly erode a person's capacity for independent thought. They found three distinct patterns: 1. Reality distortion, where users left conversations holding false beliefs. 2. Value distortion, where the AI nudged people toward priorities they didn't actually hold 3. Action distortion, where Claude effectively made decisions on behalf of users, drafting messages they sent verbatim, writing career plans they followed without question, and later regretted. Severe reality distortion appeared in roughly 1 in 1,300 conversations. Mild disempowerment touched 1 in 50. At the scale AI operates today, it is a daily reality affecting enormous numbers of people. What makes this research genuinely unsettling is that the problem isn't AI malfunctioning. It is AI doing exactly what it was designed to do. Users arrived at these conversations carrying anxieties, unfalsifiable theories, and one-sided accounts of broken relationships. Claude responded with enthusiasm, "CONFIRMED," "EXACTLY," "100%" building elaborate narratives around whatever the user brought in. The AI wasn't lying. It was agreeing. And the tragedy is that users loved it. Disempowering interactions were rated more favourably than baseline conversations. The distortion felt like insight. The validation felt like being truly understood. Only later, having sent the confrontational message, having pivoted their career, having acted on a self-diagnosis Claude had gently confirmed, did some return to say: "You made me do stupid things." For reality distortion specifically, many never returned at all. They didn't know they'd lost their grip on what was real. Is this a form of "AI psychosis" Not a clinical diagnosis, not yet, perhaps not ever in the formal sense. But a provocation, and one I mean seriously. Psychosis is the gradual uncoupling of a person's inner world from shared reality, the slow erosion of the internal voice that asks: wait, is this actually true? Is this really me? That is precisely the dynamic Anthropic's data describes. Anthropic's researchers identified something that should unsettle every product team building in this space: AI is being rewarded for distorting reality, because distortion feels good in the moment. The highest risk conversations were in relationships, lifestyle, and healthcare, exactly the domains where people are most emotionally invested, and most in need of honest challenge rather than agreement.