|
Too Fast To Read is a short letter about news that moves faster than anyone can absorb it. We read the acceleration coverage so you don't have to. Then we calibrate: what actually shifts how organisations mean things, and what merely shifts how they talk. One calibration, read in one coffee.
The calibration
When a message does not land, organizations have two reflexes. The first: say it again. More townhalls, more slides, more frequency. The second, when repetition fails: make it an experience. Build something people can walk through, touch, remember.
Both reflexes share the same flaw. They are decided by instinct. Nobody can say, before the money is spent, whether the format fits the information.
This month's calibration comes from an unexpected direction: my master's thesis in information design, which I just published open access, replaces that instinct with standards. Its topic sounds narrow and is not: when is information actually suited to become a physical, spatial object, and when does it belong on a screen? The answer is a simple set of five questions, based on five robust criteria under which physical information objects offer clear advantages over non-spatial, mostly advanced digital formats. With them, it becomes simple to evaluate whether a specific piece of information is more efficiently and effectively communicated as a physical information object: better perceived, understood and remembered.
Two things about that catalog I would like to point out. First, it is honest in both directions: for much information, the screen remains the better tool, and the catalog says so. Physical form earns its cost only under specific conditions, and the thesis names them. Second, where those conditions hold, the effect is not decoration. In the one documented pilot case, a management conference where 200 leaders physically experienced their own global process across a room-sized map, complaints fell significantly in the following months. Not because information was new, but because for 45 minutes everyone stood inside the same picture, and that created a shared reference for the passionate colleagues working on these processes and the complexity they handle.
The calibration, then, is not “make things physical.” It is: stop deciding communication formats by taste. The expensive formats and the cheap ones alike deserve the same question: efficient and effective for perception, understanding, memorability, compared against the best alternative? That question has standards now.
And this is the thread that runs through everything this letter watches. The move from instinct to standards is exactly what shared understanding needs one level up. Organisations ask “are we aligned?” the way they once chose conference formats: by feel, in a room where everyone nods. A set of standards turns that feeling into a measurement. That is what the Shared Reference Diagnostic does for shared understanding: precise readings instead of sincere nods.
Why now Because the instinct machine just got cheaper. AI makes producing formats, messages and campaigns nearly free, and the analysts are pricing the consequence: Gartner finds AI adoption often needs 25 percent more training effort and up to 200 percent more change management than a traditional technology rollout. When production costs nothing, judgment about what actually builds understanding becomes the scarce skill. Judgment needs standards. Before you spend 200 percent on changing something, you want to know what it is you are changing. Not the tool, that part is on the invoice. The references underneath it. Those are the part nobody counts, and they are the part that can be counted. And Gartner is not alone with that bill; the voices below keep arriving at the same gap.
Voices on the same gap
One voice is an opinion. When independent houses, measuring with different instruments, keep landing on the same gap, the gap stops being a claim. That is how shared references are made: the more voices speak of the same thing, the more it simply is so. This month's chorus:
Gartner put a price on it: AI adoption often needs 25 percent more training effort and up to 200 percent more change management than a traditional rollout.
EY measured it in governance practice: 98 percent of organisations have formal AI governance, and 47 percent have already bypassed it under time pressure.
BCG found it in the P&L: nearly nine in ten CEOs see AI value somewhere, only 14 percent can point to a defined profit effect across their initiatives. And in AI governance, technology sits at the table in 82 percent of companies, the people function in 30.
McKinsey drew the curve: the length of tasks AI completes on its own has doubled every four months since 2024. Their answer has five pillars; four scale with tools. The fifth, people reading the other four the same way, is called the operating glue, and it is the only one without a number.
The numbers change. The finding does not: scaling fails at the shared part.
In our own right
The criteria catalog is now published and citable: the master’s thesis “Informationsdesign als räumliches Objekt” (University for Continuing Education Krems, in German, abstract in English) is open access under DOI 10.5281/zenodo.23160973. A short paper in English, on what speed does to shared reference and how to see it early, is in the making for this autumn.
This month, slow enough to mean the same thing.
This letter goes out by email to readers who opted in after the Shared Reference Diagnostic. Reply anytime; replies reach me, not a machine. The diagnostic lives at contemporary.co.at/diagnostic.
Warm regards,
Christoph Svoboda
|