Understanding AI Systems

𝗧𝗵𝗲 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 𝗦𝘁𝗮𝗶𝗿𝗰𝗮𝘀𝗲 represents the 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲𝗱 𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 from passive AI models to fully autonomous systems. Each level builds upon the previous, creating a comprehensive framework for understanding how AI capabilities progress from basic to advanced: BASIC FOUNDATIONS: • 𝗟𝗮𝗿𝗴𝗲 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗠𝗼𝗱𝗲𝗹𝘀: The foundation of modern AI systems, providing text generation capabilities • 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀 & 𝗩𝗲𝗰𝘁𝗼𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀: Critical for semantic understanding and knowledge organization • 𝗣𝗿𝗼𝗺𝗽𝘁 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴: Optimization techniques to enhance model responses • 𝗔𝗣𝗜𝘀 & 𝗘𝘅𝘁𝗲𝗿𝗻𝗮𝗹 𝗗𝗮𝘁𝗮 𝗔𝗰𝗰𝗲𝘀𝘀: Connecting AI to external knowledge sources and services INTERMEDIATE CAPABILITIES: • 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁: Handling complex conversations and maintaining user interaction history • 𝗠𝗲𝗺𝗼𝗿𝘆 & 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹 𝗠𝗲𝗰𝗵𝗮𝗻𝗶𝘀𝗺𝘀: Short and long-term memory systems enabling persistent knowledge • 𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻 𝗖𝗮𝗹𝗹𝗶𝗻𝗴 & 𝗧𝗼𝗼𝗹 𝗨𝘀𝗲: Enabling AI to interface with external tools and perform actions • 𝗠𝘂𝗹𝘁𝗶-𝗦𝘁𝗲𝗽 𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴: Breaking down complex tasks into manageable components • 𝗔𝗴𝗲𝗻𝘁-𝗢𝗿𝗶𝗲𝗻𝘁𝗲𝗱 𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸𝘀: Specialized tools for orchestrating multiple AI components ADVANCED AUTONOMY: • 𝗠𝘂𝗹𝘁𝗶-𝗔𝗴𝗲𝗻𝘁 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻: AI systems working together with specialized roles to solve complex problems • 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀: Structured processes allowing autonomous decision-making and action • 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 & 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻-𝗠𝗮𝗸𝗶𝗻𝗴: Independent goal-setting and strategy formulation • 𝗥𝗲𝗶𝗻𝗳𝗼𝗿𝗰𝗲𝗺𝗲𝗻𝘁 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 & 𝗙𝗶𝗻𝗲-𝗧𝘂𝗻𝗶𝗻𝗴: Optimization of behavior through feedback mechanisms • 𝗦𝗲𝗹𝗳-𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗔𝗜: Systems that improve based on experience and adapt to new situations • 𝗙𝘂𝗹𝗹𝘆 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗔𝗜: End-to-end execution of real-world tasks with minimal human intervention The Strategic Implications: • 𝗖𝗼𝗺𝗽𝗲𝘁𝗶𝘁𝗶𝘃𝗲 𝗗𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁𝗶𝗮𝘁𝗶𝗼𝗻: Organizations operating at higher levels gain exponential productivity advantages • 𝗦𝗸𝗶𝗹𝗹 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁: Engineers need to master each level before effectively implementing more advanced capabilities • 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗣𝗼𝘁𝗲𝗻𝘁𝗶𝗮𝗹: Higher levels enable entirely new use cases from autonomous research to complex workflow automation • 𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲 𝗥𝗲𝗾𝘂𝗶𝗿𝗲𝗺𝗲𝗻𝘁𝘀: Advanced autonomy typically demands greater computational resources and engineering expertise The gap between organizations implementing advanced agent architectures versus those using basic LLM capabilities will define market leadership in the coming years. This progression isn't merely technical—it represents a fundamental shift in how AI delivers business value. Where does your approach to AI sit on this staircase?

AI Engineering has levels to it: – Level 1: Using AI Start by mastering the fundamentals: -- Prompt engineering (zero-shot, few-shot, chain-of-thought) -- Calling APIs (OpenAI, Anthropic, Cohere, Hugging Face) -- Understanding tokens, context windows, and parameters (temperature, top-p) With just these basics, you can already solve real problems. – Level 2: Integrating AI Move from using AI to building with it: -- Retrieval Augmented Generation (RAG) with vector databases (Pinecone, FAISS, Weaviate, Milvus) -- Embeddings and similarity search (cosine, Euclidean, dot product) -- Caching and batching for cost and latency improvements -- Agents and tool use (safe function calling, API orchestration) This is the foundation of most modern AI products. – Level 3: Engineering AI Systems Level up from prototypes to production-ready systems: -- Fine-tuning vs instruction-tuning vs RLHF (know when each applies) -- Guardrails for safety and compliance (filters, validators, adversarial testing) -- Multi-model architectures (LLMs + smaller specialized models) -- Evaluation frameworks (BLEU, ROUGE, perplexity, win-rates, human evals) Here’s where you shift from “it works” to “it works reliably.” – Level 4: Optimizing AI at Scale Finally, learn how to run AI systems efficiently and responsibly: -- Distributed inference (vLLM, Ray Serve, Hugging Face TGI) -- Managing context length and memory (chunking, summarization, attention strategies) -- Balancing cost vs performance (open-source vs proprietary tradeoffs) -- Privacy, compliance, and governance (PII redaction, SOC2, HIPAA, GDPR) At this stage, you’re not just building AI—you’re designing systems that scale in the real world. What else would you add?

The AI gave a clear diagnosis. The doctor trusted it. The only problem? The AI was wrong. A year ago, I was called in to consult for a global healthcare company. They had implemented an AI diagnostic system to help doctors analyze thousands of patient records rapidly. The promise? Faster disease detection, better healthcare. Then came the wake-up call. The AI flagged a case with a high probability of a rare autoimmune disorder. The doctor, trusting the system, recommended an aggressive treatment plan. But something felt off. When I was brought in to review, we discovered the AI had misinterpreted an MRI anomaly. The patient had an entirely different condition—one that didn’t require aggressive treatment. A near-miss that could have had serious consequences. As AI becomes more integrated into decision-making, here are three critical principles for responsible implementation: - Set Clear Boundaries Define where AI assistance ends and human decision-making begins. Establish accountability protocols to avoid blind trust. - Build Trust Gradually Start with low-risk implementations. Validate critical AI outputs with human intervention. Track and learn from every near-miss. - Keep Human Oversight AI should support experts, not replace them. Regular audits and feedback loops strengthen both efficiency and safety. At the end of the day, it’s not about choosing AI 𝘰𝘳 human expertise. It’s about building systems where both work together—responsibly. 💬 What’s your take on AI accountability? How are you building trust in it?

The AI Coding Revolution Is Here, But Are We Testing for It? As AI-assisted development reshapes how we build software, I've been thinking a lot about something that is talked about often but doesn't always get the focus it deserves: automated testing. At JPMorganChase, we're embracing AI coding tools to accelerate delivery, reduce toil, and empower our teams to focus on the work that matters, reducing cognitive load of repetitive tasks. But speed without safety is just risk in disguise. Here's what I believe every leader (and this is broader than technology) needs to consider right now: • AI writes code faster than humans can review it manually. If your testing strategy is still largely manual, you're already behind. AI-generated code can introduce subtle logic errors, security vulnerabilities, or edge-case failures that look perfectly reasonable on the surface. Automated testing is no longer a best practice, it's a non-negotiable safeguard. • Test coverage is your new quality contract. When AI is your co-developer, the test suite becomes the specification. If you can't describe expected behavior in a test, you can't trust what the AI builds. Investing in robust unit, integration, and regression testing frameworks is investing in the integrity of your entire delivery pipeline. • Shift-left testing amplifies AI's value. It doesn't slow it down. Some worry that rigorous testing will negate the speed gains from AI coding. The opposite is true. When automated tests are embedded early in the development lifecycle, AI tools can iterate faster, self-correct, and validate outputs in real time. Testing enables velocity; it doesn't constrain it. • Your teams need to evolve alongside the tools. The best teams of tomorrow won't just write code. They'll architect test strategies, evaluate AI outputs critically, and build systems that are observable and verifiable by design. We owe it to our teams to invest in this skill evolution now. At the scale we operate, serving millions of customers, the cost of a defect isn't just technical. It's trust. And trust, once broken, is hard to rebuild. AI is a force multiplier. But multiplying without a strong foundation multiplies risk just as fast as it multiplies output. Build fast. Test smarter. Ship with confidence. I'd love to hear how other leaders are thinking about quality engineering in the age of AI. What's working for your teams? #AIEngineering #SoftwareTesting

Every time I call an Uber I'm reminded of AI's biggest lie. That $100 ride used to cost $3 in 2015. We got hooked on convenience while VCs subsidized our addiction. Now I'm watching the exact same playbook unfold with AI tools and most people have no idea what's coming. I keep thinking about a conversation I had with a startup founder last week. She was bragging about how her team of 3 was now doing the work of 15 people thanks to AI tools. "We're saving $800K in salaries," she said. "It's incredible." But here's what she didn't realize: She's living in the calm before the storm. 𝗪𝗲'𝗿𝗲 𝗶𝗻 𝘁𝗵𝗲 𝗨𝗯𝗲𝗿 𝟮𝟬𝟭𝟱 𝗺𝗼𝗺𝗲𝗻𝘁 𝗼𝗳 𝗔𝗜. Remember when Uber rides cost $3 across town? When they threw promo codes at us like confetti? Venture capital was bleeding money to get us addicted to convenience. Then the subsidies stopped. That $3 ride became $25, then $100 . We were hooked, so we paid. AI is following the exact same playbook, and the signs are everywhere: → OpenAI just hired a "CEO of Applications" (hello, monetization strategy) → Claude, ChatGPT, and others are still pricing at consumer rates despite enterprise-level capability → VCs have pumped $50+ billion into AI companies that need to show returns 𝗧𝗵𝗲 𝗿𝗲𝗮𝗹𝗶𝘁𝘆 𝗰𝗵𝗲𝗰𝗸 𝗶𝘀 𝗰𝗼𝗺𝗶𝗻𝗴 𝗳𝗮𝘀𝘁. When an AI tool can genuinely replace 2 full-time employees, it won't cost $20/month forever. It'll cost what those employees cost potentially $100K+ annually. Think I'm being dramatic? Look at enterprise software pricing. Salesforce charges $300/user/month. Adobe Creative Suite went from $50/month to $600/year per license. These companies price based on value delivered, not development costs. 𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝗮𝘁 𝘀𝗺𝗮𝗿𝘁 𝗽𝗿𝗼𝗳𝗲𝘀𝘀𝗶𝗼𝗻𝗮𝗹𝘀 𝗮𝗿𝗲 𝗱𝗼𝗶𝗻𝗴 𝗿𝗶𝗴𝗵𝘁 𝗻𝗼𝘄: 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝘄𝗵𝗶𝗹𝗲 𝗶𝘁'𝘀 𝗰𝗵𝗲𝗮𝗽. That side project you've been planning? That startup idea collecting dust? This is your window. AI development costs will never be this low again. 𝗦𝗵𝗮𝗿𝗽𝗲𝗻𝗶𝗻𝗴 𝗰𝗼𝗿𝗲 𝘀𝗸𝗶𝗹𝗹𝘀. When AI becomes prohibitively expensive for daily tasks, professionals who can write, analyze, code, and strategize without $1,000/month in AI subscriptions will command premium rates. 𝗖𝗿𝗲𝗮𝘁𝗶𝗻𝗴 𝗔𝗜-𝗮𝘀𝘀𝗶𝘀𝘁𝗲𝗱 𝘄𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀 𝘁𝗵𝗮𝘁 𝗱𝗼𝗻'𝘁 𝗱𝗲𝗽𝗲𝗻𝗱 𝗼𝗻 𝗔𝗜. Learn the processes. Understand the thinking. Use AI to accelerate, not replace, your capabilities. The gold rush pricing won't last. The question is: Are you building wealth during the gold rush, or just getting addicted to cheap gold?

What’s the best way to document what AI decided—and why? Most people think AI documentation is about logging data — capturing outputs, timestamps, and probabilities. But I think that’s missing the real point. When humans make decisions, we can explain our intent. When AI makes a decision, it can’t. Because AI doesn’t have reasons — it has optimizations. And that, to me, changes everything. We don’t just need to record what the AI did. We need to explain what we taught it to optimize for. Right now, most documentation serves auditors and regulators. But the real value lies elsewhere — in understanding the alignment gap between human intent and machine logic. That gap is where trust, accountability, and learning live. Here’s how I think we can fix it: ✅ Document the goal — what was the AI trying to achieve? ✅ Note the key factors — what influenced the outcome most? ✅ Capture human intent — what assumptions shaped the model? ✅ Acknowledge uncertainty — what could still be wrong? This way, documentation becomes more than proof — it becomes memory. A shared record of how humans and machines reason together. So, what do you think — should AI documentation serve compliance, or should it serve understanding and accountability first? #AITransparency #AITrust #ResponsibleAI #AIEthics #HumanCenteredAI #AIGovernance

You know all those arguments that LLMs think like humans? Turns out it's not true 😱 In our new paper we put this to the test by checking if LLMs form concepts the same way humans do. Do LLMs truly grasp concepts and meaning analogously to humans, or is their success primarily rooted in sophisticated statistical pattern matching over vast datasets? We used classic cognitive experiments as benchmarks. What we found is surprising... 🧐 We used seminal datasets from cognitive psychology that mapped how humans actually categorize things like "birds" or "furniture" ('robin' as a typical bird). The nice thing about these datasets is that they are not crowdsourced, they're rigorous scientific benchmarks. We tested 30+ LLMs (BERT, Llama, Gemma, Qwen, etc.) using an information-theoretic framework that measures the trade-off between: - Compression (how efficiently you organize info) - Meaning preservation (how much semantic detail you keep) Finding #1: The Good News LLMs DO form broad conceptual categories that align with humans significantly above chance. Surprisingly (or not?), smaller encoder models like BERT outperformed much larger models. Scale isn't everything! Finding #2: But LLMs struggle with fine-grained semantic distinctions. They can't capture "typicality" - like knowing a robin is a more typical bird than a penguin. Their internal concept structure doesn't match human intuitions about category membership. Finding #3: The Big Difference Here's the kicker: LLMs and humans optimize for completely different things. - LLMs: Aggressive statistical compression (minimize redundancy) - Humans: Adaptive richness (preserve flexibility and context) This explains why LLMs can be simultaneously impressive AND miss obvious human-like reasoning. They're not broken - they're just optimized for pattern matching rather than the rich, contextual understanding humans use. What this means: - Current scaling might not lead to human-like understanding - We need architectures that balance compression with semantic richness - The path to AGI ( 😅 ) might require rethinking optimization objectives Our paper gives tools to measure this compression-meaning trade-off. This could guide future AI development toward more human-aligned conceptual representations. Cool to see cognitive psychology and AI research coming together! Thanks to Chen Shani, Ph.D., who did all the work and Yann LeCun and Dan Jurafsky for their guidance

AI agents are not yet safe for unsupervised use in enterprise environments The German Federal Office for Information Security (BSI) and France’s ANSSI have just released updated guidance on the secure integration of Large Language Models (LLMs). Their key message? Fully autonomous AI systems without human oversight are a security risk and should be avoided. As LLMs evolve into agentic systems capable of autonomous decision-making, the risks grow exponentially. From Prompt Injection attacks to unauthorized data access, the threats are real and increasingly sophisticated. The updated framework introduces Zero Trust principles tailored for LLMs: 1) No implicit trust: every interaction must be verified. 2) Strict authentication & least privilege access – even internal components must earn their permissions. 3) Continuous monitoring – not just outputs, but inputs must be validated and sanitized. 4) Sandboxing & session isolation – to prevent cross-session data leaks and persistent attacks. 5) Human-in-the-loop, i.e., critical decisions must remain under human control. Whether you're deploying chatbots, AI agents, or multimodal LLMs, this guidance is a must-read. It’s not just about compliance but about building trustworthy AI that respects privacy, integrity, and security. Bottom line: AI agents are not yet safe for unsupervised use in enterprise environments. If you're working with LLMs, it's time to rethink your architecture.

Use this Super Simple Post to Understand the Evolution of AI Agents in 6 Key Phases.     Often, I see confusion surrounding the development pathway from basic LLMs to fully-fledged AI Agents.   To clear the fog, I've put together a straightforward, step-by-step visualization that encapsulates the entire evolutionary journey.   Remember, this isn't merely a technical diagram, but harmoniously intertwined view of how AI systems have evolved to become increasingly capable and autonomous.   👉 Phase 1: The Foundation - Basic LLM - Simple workflow: Input (Text) → LLM → Output (Text) - Transformer-based architecture trained on vast datasets - Limited to text processing within context window - No external tools or memory capabilities   👉 Phase 2: Document Processing Capabilities - Enhanced workflow: Input (Text/Documents) → LLM → Output (Text/Documents) - Expanded context window for processing larger documents - Improved tokenization for handling structured content - Limited by static knowledge from training data   👉 Phase 3: Introduce RAGs and Tool Integration to: - Enable access to up-to-date information - Supplement LLM knowledge with external data - Improve factual accuracy and reduce hallucinations - Support specialized operations through API calls   👉 Phase 4: Integrating Memory Systems to: - Maintain context across interactions - Enable personalization based on past exchanges - Store and retrieve relevant information - Support long-running tasks and conversations   👉 Phase 5: Implement Multi-Modal Processing by: - Handling diverse input types (text, images, tables) - Generating varied output formats - Creating more comprehensive understanding - Enabling richer information exchange   👉 Phase 6: Future of AI Agent Architecture through: - Chain-of-thought processing for complex problems - Step-by-step evaluation of solutions - Dynamic tool selection based on tasks - Goal-oriented execution with self-correction   If you're looking to implement AI agents in your systems, understanding this evolutionary path is crucial.   Here are some additional tips for building AI Agents:   Start small. Don't try to build a fully autonomous agent with all capabilities at once.   Start with enhancing a basic LLM with one capability (like RAG) and then gradually add more components as you validate each integration.   Integrate thoughtfully. The more capabilities you add to your agent, the more complex the system becomes.   Monitor extensively. Track not just technical metrics but also output quality, hallucination rates, tool usage patterns, and user satisfaction to continuously refine ai agents.   Here are key capabilities to build into your architecture:   🧠 Strong Foundation LLM 🔄 Effective RAG Implementation 🛠️ Versatile Tool Use Integration 💾 Contextual Memory Systems 🖼️ Multi-Modal Processing 🔍 Self-Monitoring Capabilities 🔒 Safety Systems   Over to you: What fascinate you most about the future architecture of AI agents?

My biggest takeaways from Fei-Fei Li: 1. Just nine years ago, calling yourself an AI company was considered bad for business. Nobody believed the technology would work back in 2016. By 2017, companies started embracing the term. Today, virtually every company calls itself an AI company. 2. The modern AI revolution started with a simple but overlooked insight from Fei Fei: AI models needed large amounts of labeled data. While researchers focused on sophisticated mathematical models and algorithms, she realized the missing ingredient was data. Her team spent three years working with tens of thousands of people across more than 100 countries to label 15 million images, creating ImageNet. This dataset became the foundation for today’s AI systems. 3. The human brain’s efficiency vastly exceeds current AI systems. Humans operate on about 20 watts of power—less than any lightbulb—yet accomplish tasks that require AI systems to use massive computing resources. Current AI still can’t do things elementary school children find easy. 4. Simply scaling current approaches won’t be enough. While adding more data, computing power, and bigger models will continue advancing AI, fundamental innovations are still needed. Throughout AI history, simpler approaches combined with enormous datasets consistently outperformed sophisticated algorithms with limited data. 5. Breakthrough technologies often start as toys or fun experiments before changing the world. ChatGPT was tweeted by Sam Altman as “Here’s a cool thing we’re playing with” and became the fastest-growing product in history. What seems like play today might transform civilization tomorrow. 6. Spatial intelligence is as crucial as language for real-world applications. In emergency situations like fires or natural disasters, first responders organize rescue efforts through spatial awareness, movement coordination, and understanding physical environments—not primarily through language. This is why world models that understand three-dimensional space represent the next frontier beyond text-based chatbots. 7. Physical robots face much harder challenges than self-driving cars, which took 20 years from prototype to street deployment and still aren’t finished. Self-driving cars are metal boxes moving on flat surfaces, trying not to touch anything. Robots are three-dimensional objects moving in three-dimensional spaces, specifically trying to touch and manipulate things. This makes robotics far harder than creating chatbots. 8. Everyone has a role in AI’s future, regardless of profession. Whether you’re an artist using AI tools to tell unique stories, a farmer participating in community decisions about AI deployment, or a nurse who could benefit from AI assistance in an overworked health-care system, you can and should engage with this technology. AI should augment human dignity and agency, not replace it—which means both using AI as a tool and having a voice in how it’s governed.