Signal in the age of infinite noise

Before AI, if five analysts said the same thing, that meant something. Now if five hundred accounts say the same thing, it might just mean they all used the same tool.

What this looks like in practice

In January of this year, the prevailing view was that a direct U.S.–Iran confrontation was unlikely. The diplomatic channels were still open. The market was not pricing meaningful conflict risk. Oil was trading like nothing was coming.

The structural picture told a different story.

More than a month before the strikes began, the indicators were already pointing to a confrontation that was more likely than not. We flagged this publicly on X on January 13 while the crowd was still dismissing the risk. When the strikes hit, and oil nearly doubled, the move caught most of the market off guard. The signal was there. The crowd just was not looking at it.

The inputs we were watching were not exotic. Public statements, internal economic pressure inside Iran, and the absence of certain de–escalation patterns. Anyone with access to the open internet could see the same things. The edge was in synthesis – reading those inputs as a single converging system rather than as separate news streams. That synthesis is the hard part. The inputs are just the inputs. The bottleneck has never been technology. It has been how the technology gets used.

This is the pattern. The information was available. The tools to process it were available. What was missing was the ability to read the signal before the crowd formed around the wrong interpretation.

The scarce resource

Most people use AI to generate. Very few use it to see.

Signal is when you can look at a situation that has the entire market confused and see the structure underneath. It is when you can hold a position that every feed is telling you to abandon, and hold it anyway, because you can see something they cannot.

The challenge for most people is not generating signal themselves. It is recognizing who actually has it. Most analysis is hedged to the point of meaninglessness – strategies for avoiding accountability dressed up as analysis.

The old filter for getting past this was credentials. It no longer predicts who is seeing clearly. Plenty of the biggest calls in recent years have been missed by traditional institutions and caught by people working outside them. What matters now is whether someone is actually seeing what is happening – recognizing patterns the crowd is missing, naming what is real before it is obvious, and being right about it often enough that it holds up over time. Once you can see clearly, you start operating on a different timeline than the rest of the market.

What comes next

We are entering an era where signal is the most valuable and least understood asset in the market. The investors, builders, and allocators who figure this out first will have a structural advantage that compounds over years. The ones who keep consuming the flood without questioning it will keep agreeing with the crowd. And the crowd will keep being wrong at the moments that matter most.

Finding rooms where real signal still shows up is getting harder. Most of the venues that claim to aggregate market intelligence are just amplifying whatever the models already spit out.

Consensus 2026 in Miami is one of the few that still functions as a filter rather than an amplifier. The people who show up have skin in the game. Their disagreements are real. Their agreements were not manufactured by the same five models everyone else is using. That kind of room is getting harder to find anywhere else. Which is why I will be there – hosting a small invite–only session about what signal extraction at scale actually looks like.

The edge will not belong to whoever has the most information, the fastest tools, or the loudest platform.

It will belong to whoever can see clearly when everyone else is drowning in noise.

That is the scarcest resource in markets right now.

And it is only getting scarcer.