Web Experts AI Lab / Writing
Long-form thinking on agents, responsible AI, open technology, privacy, and the systems we build.
Browse by topic
AI leadership · 4 min read
Transformative AI can exist long before the public sees it. The decisive question is not only what technology can do, but who chooses to deploy it and under which constraints.
Read article ↗AI history · 4 min read
Conversational AI did not begin with ChatGPT. ELIZA showed in 1966 how readily people attribute understanding to a machine that reflects language back to them.
Read article ↗
Responsible AI · 4 min read
AI can interpret, choose, persuade, and act. Anyone building with it must treat ethics and oversight as part of the engineering work.
Read article ↗
Custom software · 7 min read
AI-assisted development is making purpose-built software practical for more organizations, but lower coding cost does not remove the need for architecture, security, testing, and ownership.
Read article ↗
Things we built · 4 min read
Fullmoon combines marketing experience with multiple AI systems to help local businesses identify stronger opportunities, better timing, and meaningful intent.
Read article ↗
AI policy · 5 min read
Frontier models are powerful, but fear-heavy messaging can produce regulation that misses the practical reality of how useful systems are built and supervised.
Read article ↗Web development · 4 min read
A broken premium slider became an opportunity to build a focused replacement—with AI accelerating the work and experienced review keeping it dependable.
Read article ↗
AI strategy · 7 min read
Models improve, regress, disappear, change policies, and get renamed. Durable AI systems are designed to switch providers without rebuilding the business around them.
Read article ↗
AI implementation · 6 min read
A focused six-month collaboration produced roughly a dozen internal agents and a dozen training resources grounded in the real work of building science.
Read article ↗
Custom software · 6 min read
Business owners can now express powerful system ideas in Claude, ChatGPT, Codex, Zapier, and Airtable. Bringing them to production is a different discipline.
Read article ↗
Digital marketing · 6 min read
Attention is permission, not an unlimited resource. A schedule built around volume can quietly train the audience—and the algorithm—to ignore you.
Read article ↗
AI and content · 7 min read
The obvious tells keep changing. The deeper problem is prose that is perfectly calibrated, completely frictionless, and untouched by anyone who cared.
Read article ↗Search and AI · 5 min read
Google’s shift from ranked links toward AI answers, agents, and generated interfaces changes how organizations must structure useful, credible web content.
Read article ↗
AI implementation · 5 min read
The hard part is no longer launching a chatbot. It is deciding where AI belongs inside a real workflow without creating security, quality, adoption, or cost problems.
Read article ↗
AI agents · 7 min read
The interface was always a translation layer between human intent and machine execution. Agents are beginning to remove it—and that changes the work humans need to do.
Read article ↗
Responsible AI · 8 min read
Most companies use far more AI than they realize. A useful policy begins with an honest inventory of models, hidden features, energy, data, and human oversight.
Read article ↗
Open-source AI · 5 min read
Restricting access may slow competitors, but it also slows the developers and researchers who create the next generation of useful technology.
Read article ↗
Things we built · 5 min read
An ad-free radio player became a live Bluesky-powered station with privacy-respecting analytics, no account, and thousands of human-curated tracks.
Read article ↗
Measured results · 5 min read
A redesign can produce dramatic numbers, but the lesson is not that a new page is magic. Results come from strategy, structure, useful tools, and years of accumulated judgment.
Read article ↗