You can't read the temperature in any of these. You can't pin a date to a frame, find a single day, or take a fact away with you. There is no chart axis, no caption, no number on the screen. Eighty-five years of daily global temperature data are sitting inside the work, but the work isn't trying to tell you what the data says.

For a long time I thought that was a problem.

I had built a tool that could turn the dataset into something you could watch. A series of algorithmic interpretations, each one taking the same numbers and putting them through a different lens. The animations were beautiful and they were rooted in something real. But every time I sat down to write about them, I hit the same wall. What was the point of motion without a fact attached? What was data art for if it couldn't tell you anything specific? It looked like animation without purpose, and animation without purpose felt like a problem I needed to solve.

The answer arrived sideways, and once it landed it changed how I think about data altogether.

What charts do, and what they don't

Charts are extraordinary instruments. They compress thousands of data points into a shape your eye can read in two seconds. They allow precise comparison. They let an analyst trace a trend, isolate an anomaly, defend a position. For people who already speak the language of charts, they are the best thing humans have ever invented for understanding numbers.

The phrase "for people who already speak the language of charts" is doing a lot of work in that sentence.

Climate scientists have been putting global temperature data on charts for decades. The charts are correct. The trend is undeniable. And the conversation has not moved as much as anyone hoped. Not because the charts are wrong, but because the people who needed convincing were never going to be convinced by a chart in the first place. Charts work for the audience that already trusts charts. Everyone else looked at the rising line, agreed it looked alarming, and went back to whatever they were doing.

This is not a failure of climate science. It is a failure of communication medium. The chart was the only tool in the room, and the chart was the wrong tool for the job that needed doing.

A chart's job is to deliver a fact. The job that the chart cannot do is make someone curious. Charts answer questions. They do not ask them. They reward people who came in already wanting the answer. They lose everyone else inside the first second of looking.

The job a chart can't do

What we have not had, until recently, is a serious attempt to use the same datasets as a way of starting conversations rather than ending them. Not as decoration. Not as a feel-good supplement to the real work. As a parallel mode of working with data, one whose entire purpose is to make people lean in instead of glaze over.

That is what these animations are. Live-rendered pieces, each one taking the same dataset (eighty-five years of daily global temperature observations, 31,412 days of warming) and putting it through a different algorithmic interpretation. None of them tell you what the temperature was on any given day. None of them name a year. None of them include text. You cannot take a fact away from any of them.

What you can take away is a question. Why does the colour shift like that? What is that pattern doing? What am I looking at? Those questions are the entire point. If they make you curious enough to come closer and read the plaque, the work has done its job. If they make you mention it to someone over dinner that night, the work has done it twice.

This is not a replacement for charts. It is a parallel mode that does what charts cannot do, in the same way that a photograph and a map are not in competition with each other. They are two different tools answering two different questions about the same place.

Why this matters beyond climate

The climate dataset is the case study, but the principle is bigger.

Every organisation alive right now is drowning in data and underserved by its own ability to communicate that data. Every annual report has the same row of bar charts. Every sustainability statement has the same coloured circles. Every employee briefing has the same dashboards. Every audience has the same response, which is to nod politely and forget what they were shown thirty seconds later. The form has been exhausted. This is not a failure of the people who make the dashboards, who are working with the only tool they were given. The tool is what's reached its limit.

What businesses have not tried, in any serious way, is treating their own data as something that can be made strange enough to be interesting again. Their emissions, their workforce growth over decades, their energy consumption at half-hourly resolution, their customer movements through a building, the seasonal rhythm of whatever it is they actually do. These are not boring datasets. They have been presented in boring ways. There is a difference.

Imagine an executive walking past a piece in their company's lobby that they slowly realise is showing the company's last twenty years rendered in colour and motion. They cannot read the numbers off it. They have to come closer to find out what they are looking at. Once they know, they tell the next visitor about it. And the next visitor asks a question. And the question turns into a conversation that would not have happened in front of a bar chart.

That is what data art does that data visualisation cannot. Not because data visualisation is failing at its own job, but because it is doing a different job. The visualisation tells you. The art makes you ask.

What the work actually is

These animations are the third part of a longer series. Parts 1 and 2 are static, large-format compositions from the same dataset. Part 3 is the moving version. None of the animations are pre-rendered. None are AI-generated. Each frame is computed from the data in real time, through software I built specifically for this practice over the last few months.

I built it that way deliberately. The point of the work is the specificity of the source. Every aesthetic decision is one I made by hand, parameter by parameter, because the dataset demanded a specific interpretation rather than a general one. A piece like this has to be tied to something real and irreducible. Otherwise it is decoration, and decoration does not start conversations.

The animations are in Part 3 of Eighty Five Seasons.

What I'm actually curious about

If you work with data in any field, climate or finance or healthcare or sport or whatever else, I would be curious whether the same logic applies in your own world. Where are charts winning the argument they are meant to win? Where are they losing? And what happens to the conversation when you stop trying to deliver the fact and start trying to start the conversation?

The honest answer is that I do not know yet. This is the question the work is asking, not a position I have arrived at. If anything in the above resonates, or if you want to push back on any of it, I would genuinely like to hear it.

Mac