Articles

🦾 My OpenClaw team's latest members; Ogilvy (marketing and design) and Vignelli (HTML, CSS, font expert) are really knocking it out of the park. At dinner with my wife on Friday night I had her look at the recommendations for the three color schemes and two font types for our...

🦾 I feel that Max, my OpenClaw AI Chief of Staff, is really beginning to earn his tokens. 🦾🦾🦾

What makes AI "Agentic"? Autonomous and time based execution are the soul of Agentic AI. Does OpenClaw manage time efficiently with your LLM tokens? Click the article below to take a trip down the Rabbit Hole as I share some experience that increased throughput and lowered costs.

The White Rabbit is the only character in Wonderland with somewhere to be — which makes him the perfect metaphor for an AI agent. Ted traces the evolution from chatbot to agentic AI, and the technical rabbit hole of building a cron system that doesn't burn tokens checking an empty inbox.

Just published: Down Alice's AI Rabbit Hole of Time — teaching AI to be agentic. The White Rabbit is the most underrated character in Wonderland. The only one with a schedule. The only one who acts without being asked. That's an AI agent. 🧵👇 1/10

Saturday my AI team was quite busy. Ken set up their local test environment using Ansible. Then James refactored over 20,000 lines of Java code plus some new features. Watts created over 40 JUnit and Playwright tests. Ken then deployed it all to the cloud via Ansible. All done via Telegram. How we work has dramatically changed.

🦾🦾🦾 Yesterday was a big day for my 'team' of AI agents. James changed 20,500 lines of Java code. Watts created 31 separate test suites in Playwright and JUnit. Ken did several deployments to both the local test and AWS prod environments with Ansible. I managed all of it from my phone.

It's slightly embarrassing when you explain the work team you have been texting with — are AI agents, despite their common names and roles… 🦾🦾🦾

Changed 20,000 lines of Java code today and got a great workout in as well as went to dinner with my wife. Thank you OpenClaw. 🦾

I have been running my AI Agent(s) for a little over a week now. I have had it solve real world business problems (OK some were part of my hobby — but it includes true AI operations management). I then went on to create a full team of AI agents that work together, share information, communicate in the background autonomously and help improve the product I built.

Just posted on LinkedIn about building my AI agent team. I've been running this AI team for a week. Here's what actually works. A thread: 🧵👇

Will Anthropic blink? Will safety prevail and dive their $380B valuation?

Is Anthropic's safety-first strategy a brilliant competitive moat or a valuation that logic can't support? Ted examines the Pentagon ultimatum, the $30B Series G, and what it means when the company that walked out of OpenAI over safety concerns faces pressure to compromise it.

Over the past two decades, I've worked across every major phase of digital discovery — from the rise of search to its reinvention in the AI era. As CTO of CNET, head of AOL Search and MapQuest, and later global head of MSN — I was inside the search wars. Today, we're entering the next great shift: from search to AI chat.

Full long-form article on X — from CTO of CNET to global head of MSN, Ted was inside the search wars. Now he examines the next great shift: from search to AI chat, and whether the incumbents' moats still hold.

Does your AI assistant schedule and negotiate meetings on your calendar? Are you running a fully autonomous AI agent at home or work? Let me share my experience with my AI agent "team" — and some key takeaways if you're even remotely interested in this topic or considering "employing" one (or more).

I hired an engineer to work on my hobby projects for $20 per month. His name is Claude. He already completed tasks that I had put off for over a year. Complete upgrades, feature requests, bug fixes, and some platform migrations. Claude Code on Opus 4.6 is a game changer.

In 2018, Ted received the Washington DC Region SCCA Regional Executive Award for building and operating marrspoints.com — the automation project that became the proof of concept for Agentilis.

A deep dive into the data analytics layer Ted built on top of marrspoints.com — race class participation trends, driver counts, and multi-year comparisons. Data engineering before it was called that.

The 2016 origin story of marrspoints.com — how Ted combined his engineering background and passion for racing to build a cloud-based points scoring system on AWS. The project that started everything.

MARRS STL class champion, 4th in SM, Ironman award winner, and the Sunoco Hard Charger Award at the Runoffs — advancing from 48th to 14th at the SCCA National Championships. The person behind the technology.

Race car drivers generate about 15-20GB of video files over a three day weekend. I have been saving mine for years on a Synology NAS on my home network. I may be the last person to learn that I can now browse them securely anywhere on earth from my phone with Tailscale.

A cheerful deep dive into replacing drives in Synology NAS units while they're still running. 30TB of RAID-5 storage, hot-swap procedures, and the kind of home infrastructure obsession that built a career.

Ted on the unexpected fitness benefits of AirPods — and why the intersection of consumer technology and personal health is more interesting than it sounds. Self-deprecating, charming, and very Ted.

Written when Ted joined block.one and relocated to Virginia — a reflection on how digital infrastructure has transformed moving. From shipping CDs to cloud-everything.

The 2016 launch post for cahall-labs.com — a media executive dusting off his personal tech writing roots. Traces the through-line from LiveJournal in 2004 to building in the cloud in 2016.