// writing & talks
Things I've been writing.
filter:
2026
- 2026.06 Anthropic takes the lead with Fable short →
- 2026.06 Building the factory that builds software essay →
- 2026.05 The model is the brain, the harness is the body essay →
- 2026.05 Rabbit and turtle essay →
- 2026.05 AI in search — embeddings is not a strategy essay →
- 2026.05 The end of repetitive tasks short →
- 2026.05 Gemini 3.5 Flash — the secret short →
- 2026.05 Is the finetuning era over? short →
- 2026.05 Harness engineering is the new prompt engineering short →
- 2026.05 Open weights update short →
- 2026.05 Mythos and Firefox short →
- 2026.05 LinkedIn hallucination rates short →
- 2026.04 The AI disruption matrix short →
- 2026.04 Clawdbot in 200 lines of code essay →
- 2026.04 Let's talk about clawdbot short →
- 2026.03 The year that Google strikes back short →
- 2026.03 The missing ingredient for AGI short →
- 2026.03 The future of work short →
- 2026.03 My take on Gemini 3 short →
- 2026.02 What makes companies fail short →
- 2026.02 Software engineers are obsolete, or are they? short →
- 2026.02 Have you read the AI 2027 analysis? short →
- 2026.01 The way we work is about to change short →
- 2026.01 Here is how we use AI at Mantis short →
2025
- 2025.12 Here is the issue with AI doing actual work short →
- 2025.12 Back to trillion param models short →
- 2025.11 Are we beginning to use AI too much? short →
- 2025.10 GPT5 is amazing, or is it? short →
- 2025.09 From AI tools to AI infrastructure short →
- 2025.09 Agentic RAG is the new oil, don't miss out! short →
- 2025.08 AgenticAI is everywhere — but what is an agent? short →
- 2025.08 How to know you're sitting on valuable insights short →
- 2025.08 Claude sub-agents short →
- 2025.07 Stop using AI as a magic wand short →
- 2025.07 Learn by not using AI or tools short →
- 2025.07 What's going on with MCP? short →
- 2025.07 Upskilling a software engineering team short →
- 2025.06 Will AI make your role obsolete? short →
- 2025.06 RAG meets knowledge bases talk →
- 2025.06 The power lies in who is using the tools, and how short →
- 2025.06 Efficiency short →
- 2025.06 AI as a productivity booster short →
- 2025.05 What should you expect from AI in the near term? short →
- 2025.05 What happened with GPT4.5? short →
- 2025.05 LLMs getting lost in multi-turn conversation short →
- 2025.04 Here is how R1 is trained short →
- 2025.04 What is going on with DeepSeek? short →
- 2025.04 TutorCopilot short →
- 2025.04 Hebbia short →
- 2025.03 AI to find patients in clinical trials short →
- 2025.03 OCR short →
- 2025.03 The year of agents short →
- 2025.02 Static embeddings short →
- 2025.01 LLM world model short →
- 2025.01 ModernBERT short →
- 2025.01 Open models short →
- 2025.01 Building agents short →
2024
- 2024.12 Pydantic AI short →
- 2024.12 RAG metrics short →
- 2024.12 Write about AI and todos short →
- 2024.12 Cerebras and Groq speeds short →
- 2024.12 My take on Reflection and o1 short →
- 2024.11 Docling short →
- 2024.11 Extract financial data short →
- 2024.10 Disrupting transformers short →
- 2024.10 Out-of-domain embeddings short →
- 2024.10 Audio from sources short →
- 2024.10 A tour of LLMs talk →
- 2024.10 Open vs closed source short →
- 2024.10 AI thinking short →
- 2024.10 GLiNER short →
- 2024.10 Contextual retrieval and late chunking short →
- 2024.10 AI producing research ideas short →
- 2024.06 Extreme multilabel classification talk →
- 2024.06 Interpretability of LLMs talk →
- 2024.01 The year of ChatGPT short →
2023
- 2023.10 Eight examples is all you need essay →
- 2023.06 Prototyping with Large Language Models short →
2022
- 2022.09 Reproducible data science essay →
- 2022.07 MLOps with SageMaker — Part II essay →
- 2022.06 MLOps with SageMaker — Part I essay →