There has been a wave of AI releases recently. New models, improved performance, better benchmarks. Each step is meaningful, and the pace of progress is real. But when you start building real systems with these tools, a different picture emerges. Most failures are not happening at the model level. They are happening in how systems […]
Human-in-the-loop is often described as a requirement for AI systems. Add a human for review.Add approval steps.Add checkpoints. On paper, that makes sense. In practice, it often fails. Not because the idea is wrong. Because the experience around it is. Most systems treat human interaction as an interruption. A modal.A notification.A separate screen. Something that […]
A lot of AI interfaces today are still built around the same patterns we’ve used for years. Pages.Forms.Buttons. Even when AI is involved, the experience often comes down to entering a prompt and waiting for output. That works, but it misses something important. It assumes the user knows what to do next. In practice, that’s […]
There’s a growing belief that AI is going to simplify software development. Write some prompts. Generate some code. Move faster. On the surface, that’s happening. But underneath, something else is going on. We’re not reducing complexity.We’re increasing it. More code is being created.More systems are being stitched together.More moving parts are showing up in places […]
There is a lot of conversation right now about how quickly AI can generate software. What is discussed less often is how products actually make it to market. AI can generate a surprising amount of functionality. It can accelerate development in ways that were difficult to imagine only a few years ago. But generating code […]
This week reinforced something important. The race in AI is no longer just about models. It’s about control. OpenAI expanding capabilities.Anthropic releasing remote control features.Perplexity launching computer use.Nvidia pushing agentic tooling.China advancing new foundation models. But another layer emerged. Alignment. OpenAI agreeing to work with the Department of Defense.Anthropic choosing a different path.Signals that federal […]
Governance around AI agents is becoming a louder topic. That’s not a reactionary trend. It’s a predictable one. When agents move from generating text to taking action inside real systems, governance stops being theoretical. Most organizations approach governance as a policy problem.Write the document. Define the rules. Approve the tool. That’s not enough. Agent governance […]
Last week was full of impressive AI headlines. New models from OpenAI and Anthropic. New benchmarks. Even stories about AI writing compilers that can run games like Doom. That level of capability is remarkable. The progress is real. But most organizations don’t struggle with what’s possible.They struggle with what’s usable. A capability without integration is […]
Winter has a way of exposing weak systems. Pipes freeze.Roads fail.Schedules fall apart. It’s rarely because someone didn’t work hard enough.It’s because the system assumed ideal conditions. Software systems behave the same way. Everything looks fine when volume is low, timelines are loose, and the same people are always around to remember how things work. […]
The trophy is easy to see. The work that earned it isn’t. When you watch a championship game played in brutal conditions, the lesson is obvious. The teams that survive aren’t the flashiest. They’re the ones that can execute the fundamentals when everything else breaks down. The 2025 Patriots are a good example of a […]
