Analysis Is a Commodity. Situation Expertise Isn't.
The analytical work that consultants and knowledge workers charge for — structuring a problem, sizing a market, summarizing a report — is being repriced by AI toward zero. What isn't being repriced is knowing what happened in your client's industry last week. That knowledge can't be generated on demand; it can only be accumulated. This piece is about why that distinction now decides who gets paid.
Two kinds of expertise
A professional's value splits into two parts. One is method: knowing how to structure a problem, run the analysis, build the model, write the deliverable. Call it outcome expertise — it's portable, you can apply it to any client. The other is context: knowing a particular market as it stands this week — who moved, what changed, what practitioners are quietly arguing about. Call it situation expertise — it isn't portable at all, and that's exactly its value.
AI has collapsed the price of the first kind. It cannot produce the second — not because models are weak, but because this knowledge isn't derived, it's accumulated. That asymmetry is what this piece is about.
What commoditization actually looks like
"Commodity" here is the economic term: a good available to everyone at near-identical quality, where nobody pays a premium. That's what has happened to baseline analysis.
I've watched it happen in my own work. Lead research used to be a manual job — before any outreach, someone on the team spent hours researching each prospect and hunting for contacts. Today systems do that across open sources continuously; nobody would price that work as expertise anymore. The same repricing is running through market overviews, competitive scans, report summaries, first-draft strategy decks. The floor rose. A client with a chatbot can get a passable version of all of it in minutes — and the client knows that.
Everyone says the AI era rewards "taste," and it's usually said about art and content. I think it applies with full force to business: in consulting, in development, in any niche, taste is a trained eye — the accumulated exposure of having watched a market move, week after week, until you can tell a real shift from noise. AI averages: it hands you the consensus of everything it has read. The value moves to the person who can see where the market has already moved past that consensus — and that only comes from watching.
There's a second effect people miss: the same tools that speed you up speed the whole industry up. More gets announced, shipped, regulated, and abandoned per quarter than before. So you need more analysis, faster, just to keep pace. Commoditization turned the analytical workload into infrastructure.
What doesn't commoditize
A general model knows the world in average. It doesn't know that at your client's market, this specific week, a regulator moved, a competitor's launch landed badly, and practitioners are quietly skeptical about the tool everyone's announcing. That knowledge has three properties that resist commoditization: it's narrow, it's fresh, and it decays — which means it has to be maintained, not generated.
Here's what that looks like in practice. I sit in meetings where we discuss replacing a proprietary model with open source. I'm current on the open-source model market — releases, benchmarks, licensing shifts — and, just as important, on what developers actually think of these systems, because I read the practitioners, not just the announcements. So when the room debates an agent framework, I already know the market has open questions about it. And when someone raises a security concern, I sometimes already know that this particular concern hasn't held up. None of that is deep analysis — it's surface domain awareness. And it changed the conversation.
That's situation expertise: not being smarter in the room — being current in the room.
The mirror case: I work with genuinely brilliant people who don't track this space. They don't look worse for it — but they need someone who brings the context, and that someone is either a person or a system. Smart plus current beats smart alone, every time, and "current" is now the scarce half.
The two layers — and being honest about them
I should be precise about what tracking gives you, because I spent years in research and I know its limits. The most valuable part of every study I ran — desk research or qualitative — was always the conversations with decision-makers. What a market leader tells you across a table is regularly different from what the public record says. No digest replaces that, mine included.
So think in two layers. The first layer is what happened: releases, filings, deals, regulatory moves, the industry's public conversation. This layer is trackable, wide, and now automatable — a system can watch twenty sources and hand you the delta every morning. The second layer is what it means: the practitioner's read, the skeptical engineer's blog post, the thing a decision-maker says off the record. That layer you build by going deeper on what the first layer surfaces — my own routine is exactly that: the digest gives me the background fast, and when something matters, I go read the practitioners I trust and, when I can, talk to the people involved.
And the two layers feed each other. Daily exposure to layer one is what builds the trained eye — the context that lets you ask practitioners and decision-makers the right questions instead of generic ones. You can't go deep on a market you haven't been watching.
The mistake is doing layer one by hand — burning hours on coverage a system does better — and then having no time left for layer two, which is where your differentiation actually lives. If your first layer is still manual, start with the setup in How to Keep Up With the News — one scheduled arrival, tracked wide, read narrow.
What I'd say to a professional worried about being left behind
The fear I hear most often isn't really "AI will take my job." It's arriving at the market a step late — the industry's own phrase for it, "left behind," comes up constantly. My answer: don't fall out of tempo. AI accelerates your work, but it accelerates the whole world at the same time — so the only stable strategy is to accelerate yourself where acceleration is available. Analysis: let the tools do the floor, keep the judgment. Currency: automate the watching, keep the reading that trains the eye. What can't be accelerated — relationships, judgment, the conversations no system attends — is where your remaining hours should go.
Analysis is becoming a commodity. Knowing what happened in your client's industry last week is not. Build your week around the second sentence.
The "what happened" layer is what Dailyn does: describe your clients' industries once, get one daily brief. Free to start. The "what it means" layer is still yours.