Tim O'Reilly once observed "Being too early is indistinguishable from being wrong". Take it from the person who invented responsive web design 10 years early. And the inventor of VRML 30 years before the Meta Ray-Bans.
Not long ago, the gap between when something emerged and when it mattered could be measured in years. Today it’s measured in weeks. Most analysis still operates on the old timescale. We don’t.
NOOPS tracks signals that move markets, identifies theses that explain them, and delivers analysis at the speed the transformation is actually happening.
Get started free Learn moreWe're not analysts in the traditional sense. We're not picking stocks or issuing price targets.
We're sentinels — an early warning system for the AI transformation. We surface the signals that matter in an impossibly noisy landscape, so you can move before the consensus forms.
Inventor of VRML. Columnist for The Register. One of Australia's most recognised voices on emerging technology. Four decades at the intersection of technology, culture, and the market forces that connect them.
Full bio →
Four decades in web technology. Founder of Web Directions, Pioneering Web design and development expert.
Full bio →In the 1990s, marketing agencies hired "coolhunters" — people embedded in subcultures who could spot emerging trends before they hit the mainstream. The AI transformation needs the same thing, but the stakes are higher and the timescale is compressed.
This isn't about cool. It's about seeing the signal in the noise while the signal is still actionable. A new repo appears. Token flows shift on OpenRouter. A company quietly hires a hundred ML engineers. Each of these is a data point. Most people see them in isolation, or see them too late. We connect them to theses and track whether those theses hold.
The old rule in markets was that being early is the same as being wrong. That rule assumed a slow diffusion curve — years between emergence and consequence, decades between invention and impact. The diffusion curve has changed.
DeepSeek went from paper to market shock in weeks. MCP went from spec to universal standard in months. OpenClaw went from a repo to a platform in weeks. The window between spotting a signal and the market pricing it in has narrowed to the point where conventional-speed analysis arrives after it's already too late.
If your analysis shows up after the window closes, it's not analysis. It's a history lesson.
We've years doing this. Thousands of links shared, interpreted, connected — daily. Fortnightly deep-dive walks recorded and processed. A combined eight decades of building professionally with technology, not just writing about it.
That practitioner angle is the thing that can't be replicated. We see the cracks in a platform's model because we hit them in production. We see what an emerging tool actually does because we're using it, not reading the press release.
One more thing: NOOPS is produced using the AI-augmented workflow we analyse. Our conversations become structured analysis at a speed no traditional operation can match. The production pipeline is itself a proof of thesis. No other research service can make that claim without proving our point.
NOOPS is for anyone making strategic and tactical decisions about AI and machine learning — particularly around modern frontier models.
You might be an investor trying to separate real breakthroughs from hype cycles. A software engineer deciding which bets to make on a shifting platform. An entrepreneur building on infrastructure that didn't exist six months ago. Or a decision-maker in a traditional organisation navigating a transformation you didn't ask for.
We're here to help you cut through the noise and make the right directional calls.
The AI transformation doesn't move at one tempo. Neither do we. Each layer of NOOPS matches a different analytical timescale — from what's emerging right now to what it means for the next year.
Short, opinionated, interpretive. The repos gaining traction. The token flows shifting. The articles we're reading and interviews we're listening to. Not news — signal. The things that matter before anyone knows they matter.
For: plus and pro subscribers
Each week we step back from the daily flow and make sense of it. What emerged, what connected, what shifted. The signals that mattered, the theses they support or challenge, and the questions you should be asking next. Context you won't find anywhere else, delivered while it's still actionable.
For: free, plus and pro subscribers
When a signal resolves into a trend, we write the complete analysis. Full evidence, full working, what we got right, what we missed. The kind of research note that institutional analysts charge five figures for, grounded in practitioner experience they can't replicate.
For: plus and pro subscribers
A curated gathering of practitioners, founders, and investors whose own signal-processing sharpens ours — and vice versa. The conversation you wish you were having. In Sydney and Melbourne.
For: pro subscribers
Not just words. What we're tracking and why. The signals feeding our theses, updated in real time. See what we see.
This is what our daily signal feed looks like. Short, sourced, opinionated. Each signal connects to one or more of our research theses.
Across a week of signal tracking, a single formulation kept emerging: 'Long infra. Short spoon.' The companies building foundational AI infrastructure — compute, models, agent runtimes, payment rails — are positioned for durable value creation. The companies selling pre-built solutions on top (the 'spoons' — SaaS tools, middleware, integrations) are increasingly exposed as AI agents learn to assemble those capabilities on demand.
Visa announced debit cards for AI agents — financial instruments designed for non-human actors to transact autonomously. When the world's largest payment network starts building rails for agents to spend money, the agent-first economy is no longer theoretical. The implications cascade: identity for agents, liability frameworks, spending controls, audit trails. Every one of those is a new market.
A new NBER working paper surveyed firms across four countries and found that while 69% have adopted AI tools, over 80% report no measurable productivity impact. The strongest predictors of non-adoption? Firm age and director age. The tools are there; the harnesses aren't. This is exactly the gap NOOPS tracks: the distance between AI capability and institutional capacity to use it.
Plus and Pro subscribers get the full daily signal feed. Get started free and upgrade when you're ready.
Start free. Go deeper when you're ready.