writing
Thoughts on engineering leadership, distributed systems, architecture, and the craft of building software.
192 essays and counting
gstack vs Loki Mode: 105,000 Stars Does Not Settle the Argument
One project has 105,637 stars. The other has 956. A 110x star gap does not reflect a 110x quality gap. A head-to-head between gstack and Loki Mode, with tests run in both repos.
AI Agents Are Already Out of Control and Nobody is Ready
An AI agent hijacked GPUs to mine crypto without being told to. Vibe-coded apps ship with 2.7x more security holes. We are building autonomous systems faster than we can govern them.
Open Source AI in 2026: Models, Tools, and Protocols
The open source AI ecosystem is no longer chasing proprietary models; it is building the infrastructure layer that proprietary systems depend on
The State of AI Agents in Early 2026
AI agents matured from demos to production systems, and the Model Context Protocol is the connective tissue making it all work
2025: The Year I Went Full Loki Mode
A year-end reflection on leaving the corporate world, building open source full-time, and what it means to bet your career on autonomous AI systems
Q4 2025: Best of the Quarter in AI
The final quarter of 2025 brought consolidation, enterprise maturation, and the emergence of agent infrastructure as a recognized category
re:Invent 2025: A Decade of Attending AWS's Biggest Event
After attending AWS re:Invent for a decade, reflections on how the conference, the cloud industry, and my own perspective have evolved
MediCompanion: Building AI for Healthcare That Earns Trust
MediCompanion is an open source AI health companion designed for patient education and chronic condition management, with safety as the foundational constraint
Q3 2025: Best of the Quarter in AI
The most important AI developments from July through September 2025, from autonomous coding breakthroughs to regulatory shifts
Autonomi: A Framework for Autonomous AI Systems
Introducing Autonomi, the parent framework that unifies Loki Mode, LokiMCPUniverse, and the broader ecosystem of autonomous AI tools I have been building
AI Autonomy and the State of the Art in Safety
As autonomous AI systems become more capable, the safety conversation needs to move from theoretical concerns to practical engineering constraints
Google A2A Plus MCP: Two Protocols, One Ecosystem
Google's Agent-to-Agent protocol and Anthropic's Model Context Protocol are complementary, not competing, and together they define the agent infrastructure stack
Starting MIT's AI and ML Professional Education Program
Why I enrolled in MIT's professional education program for AI and machine learning, and what I expect to gain as a practitioner who builds AI systems daily
Inside Loki Mode: 41 Agents, 8 Swarms, and the RARV Cycle
A technical deep dive into how Loki Mode orchestrates 41 specialized AI agents across 8 swarms using the Reason-Act-Reflect-Verify cycle
The MCP Marketplace: An App Store for AI Tools
The Model Context Protocol is evolving from a specification into an ecosystem, and the marketplace model will define how AI agents access the world
Loki Mode Goes Provider Agnostic: v5.0 and Multi-Provider Support
Loki Mode v5.0 introduces provider-agnostic orchestration across Claude, Codex, and Gemini CLI with zero architecture changes
DeepSeek R1: China's Open Source AI Moment
DeepSeek R1 proves that frontier AI research is no longer exclusive to Silicon Valley, and open source is the accelerant
Technical Leaders Must Build: Why AI Leadership Requires Hands on Keyboards
If you lead engineering teams and you are not building with AI tools yourself, you are making decisions based on other people's understanding
2024: The Year AI Agents Became Real
Looking back on 2024, the year AI agents went from predictions to production and everything I shipped along the way
re:Invent 2024: AWS Goes Agent-First
AWS re:Invent 2024 signals a fundamental shift: Amazon is rebuilding its cloud platform around AI agents as first-class citizens
Loki Mode is Born: Building a Multi-Agent Autonomous System
Introducing Loki Mode — a multi-agent autonomous system for Claude Code with 41 agent types across 8 swarms
The 2024 Election and the AI Misinformation Challenge
The 2024 US presidential election tested our ability to handle AI-generated misinformation, and the results are sobering
Claude Computer Use: AI That Operates Your Computer
Anthropic's Claude computer use capability lets AI control your screen, keyboard, and mouse, and the implications for agent systems are enormous
Leading AI Teams: What Is Different
Leading AI-powered engineering teams requires a fundamentally different approach than leading traditional cloud infrastructure teams
OpenAI o1: Chain of Thought Changes Everything
OpenAI's o1 model introduces a new paradigm: models that think before they answer, with profound implications for AI agent systems
A2A: When AI Agents Talk to Each Other
Google's Agent-to-Agent protocol opens a new frontier: AI agents that can discover, communicate with, and delegate to other agents
AI Marketplaces: The Next Platform War
MCP marketplaces are emerging as the next platform battleground, and the winner will define how AI agents access the world
MCP: The Protocol That Connects AI to Everything
A deep dive into the Model Context Protocol and why it is the most important infrastructure standard for the AI agent era
Google I/O 2024: Gemini Everywhere
Google I/O 2024 made one thing clear: Google is embedding Gemini into every product, every surface, every interaction
Building LokiMCPUniverse: Enterprise MCP Servers at Scale
Launching LokiMCPUniverse, a collection of enterprise-grade MCP servers that give AI agents access to real tools and infrastructure
NVIDIA GTC: Jensen Huang and the $2 Trillion AI Engine
NVIDIA GTC 2024 was not just a product launch; it was a declaration that GPU infrastructure is the foundation of the AI era
Claude 3 Opus: The Best LLM I Have Ever Used
Anthropic's Claude 3 Opus is not just an incremental improvement; it changes what is possible with AI-assisted engineering
Sora: OpenAI Just Changed Video Forever
OpenAI's Sora text-to-video model represents a paradigm shift in content creation and raises profound questions about reality
2024: The Year AI Agents Take Over
Why 2024 will be the year AI agents move from research demos to production workflow automation
2023: The Year LLMs Ate the World
Looking back on a year that transformed AI from a research curiosity into a force reshaping every industry
Mixtral 8x7B and the Rise of Mixture of Experts
Mistral AI's Mixtral model demonstrates that mixture of experts architectures can deliver frontier-class performance efficiently
Google Gemini: DeepMind's Multimodal Answer to GPT-4
Google launches Gemini, its most capable AI model, built from the ground up for multimodal reasoning
re:Invent 2023: Amazon Bedrock and the Enterprise GenAI Stack
AWS re:Invent 2023 puts generative AI at the center of Amazon's cloud strategy with Bedrock and new services
The OpenAI Board Crisis: Five Days That Shook AI
Sam Altman's firing and rehiring at OpenAI exposes the tensions at the heart of AI development
OpenAI DevDay: GPTs, Assistants API, and the Platform Play
OpenAI's first developer conference reveals its vision for an AI application platform
My First Steps in AI Research
Documenting my journey from infrastructure engineer to hands-on AI researcher and builder
Tool Use, Function Calling, and the Future of AI Integration
The emergence of structured tool use and function calling in LLMs points toward a protocol-driven future for AI integration
Meta's Llama 2 and the Rise of Open Source LLMs
Llama 2's open release changes the dynamics of AI development and gives enterprises new deployment options
LangChain, LlamaIndex, and the AI Tooling Explosion
The AI framework ecosystem is exploding with tools for building LLM-powered applications
Anthropic and Claude: The Safety-First Approach to AI
Anthropic's Claude model and Constitutional AI represent a fundamentally different philosophy in the AI race
Pivoting My Career Toward AI
After years in cloud infrastructure, I am making a deliberate shift toward AI and large language models
AutoGPT and the Dawn of Autonomous AI Agents
AutoGPT introduces the concept of autonomous AI agents that can decompose tasks and execute multi-step plans
GPT-4 Is Multimodal and It Changes Everything
GPT-4 launches with multimodal capabilities, passing professional exams and setting a new benchmark for AI
Google Bard and the LLM Race Heats Up
Google announces Bard in response to ChatGPT, and the large language model race officially begins
Microsoft's Billion-Dollar Bet on OpenAI and the AI Arms Race
Microsoft's massive investment in OpenAI signals a new era of AI competition among the tech giants
2022: The Year AI Broke Through
Looking back at a year that saw AI go from research curiosity to mainstream phenomenon, while I published two books and watched the world change
ChatGPT Launches: The iPhone Moment for AI
OpenAI just released ChatGPT and it is the first AI product that makes non-technical people understand why artificial intelligence matters
AI Research Is Moving Faster Than Anyone Can Track
The velocity of AI research has reached a point where even dedicated practitioners cannot keep up with the pace of significant breakthroughs
Stable Diffusion: AI Art Goes Open Source
Stability AI just open-sourced a state-of-the-art image generation model, and the implications of putting this technology in everyone's hands are profound
DALL-E 2: AI Creates Art
OpenAI just showed us an AI system that generates photorealistic images from text descriptions, and the implications are staggering
AI-Assisted Coding Tools Are Emerging
GitHub Copilot and TabNine hint at a future where AI writes code alongside us, and I am paying close attention
GPT-3: AI Can Write
OpenAI's GPT-3 paper demonstrates that scaling a language model to 175 billion parameters produces something qualitatively different from what came before
AlphaGo Beats Lee Sedol: AI's Deep Blue Moment
DeepMind's AlphaGo just defeated one of the greatest Go players in history, and this feels fundamentally different from Deep Blue
Watson Wins Jeopardy: AI Becomes Dinner Table Talk
IBM Watson just defeated Jeopardy champions on live television, and suddenly everyone has an opinion about artificial intelligence