Ask a room to describe technology. Then count the bruises.
Nobody in the room is a pessimist. Ask them individually and most will tell you they are optimistic about where this is going. Then ask them for one word, go round the table, and write down what you hear. The wall that builds is not what any of them would have predicted — and the gap between the two is the most useful thing you will learn all day.
Guess first — commit before you collect
Before a single word is spoken: what share of them will be negative?
You are about to ask this room for single words describing technology and the last decade of change. Commit to a number now, out loud, in front of everyone. Then go and collect the words and find out.
Collect it
Now go round the room
One word each. Type them in as they are called out — the wall builds on the screen behind you. Words are sorted automatically, and the sorting is a guess, not a verdict: click any word to change it. Words the list doesn't recognise arrive dashed and neutral, waiting for you to decide.
Enter adds the word. Click a word to cycle it negative → neutral → positive. Shift-click removes it. Nothing leaves this browser.
Why the wall leans
The room is not pessimistic. The room is calibrated.
The people who gave you those words are not gloomy. Ask them privately whether they expect their own lives to be better in ten years and most will say yes. What you have measured is not their forecast — it is which parts of their forecast arrive first, loudest, and with the most detail attached. Bad news has four structural advantages, and none of them is about the news being true.
Bad is stronger than good
Across a very wide range of psychology — impressions, relationships, learning, feedback — negative events reliably outweigh comparable positive ones. It is one of the most replicated patterns in the field, and the usual summary is blunt: bad is stronger than good.
It is not a flaw. An ancestor who ignored one good berry ate less. One ignored predator ended the line. Asymmetric stakes produce asymmetric attention.
Losses weigh roughly twice as much
In Kahneman and Tversky's work, a loss is felt about twice as strongly as an equivalent gain. Which means an accurate ledger — this change costs us X and returns us 2X — still feels like a wash to the person reading it.
Anyone proposing change is arguing uphill against a factor of two, and usually doesn't know it.
Costs are specific and now; benefits are vague and later
The cost of a new system has a number, an owner, a date, and someone whose job changes. The benefit is a percentage, applying to a category of person, starting sometime after the next budget cycle.
Concrete beats abstract in every argument. The pessimistic case is not better reasoned; it is better staffed with detail.
The negative version travels further
Attention is bought and sold, and the market has a preference. A 2023 study in Nature Human Behaviour tested headline variants at scale and found negative words raised click-through while positive words lowered it.
So the same event reaches the room in its most alarming available phrasing — not because anyone lied, but because that phrasing out-competed the others on the way there.
Put the four together and you get a room that is individually hopeful and collectively grim, with no one in it having said anything untrue. That is worth saying out loud before you show anyone a strategy: the wall is a measure of what reached them, not a measure of what is happening.
Where this bites
The same tilt, wearing work clothes
Once you have seen the wall you start seeing the tilt everywhere, usually dressed as rigour. In each of these the negative reading gets treated as the neutral one, and the burden of proof lands entirely on the person hoping.
The risk log has no upside column
Every organisation maintains a careful list of things that could go wrong and no list at all of things that could go right, or of what it costs to keep not doing them. The asymmetry is in the furniture.
Caution reads as seniority
"I have concerns" sounds experienced. "This could be very good" sounds naive. Rooms reward the first and discount the second, so the measured optimists go quiet and the record shows a consensus that never existed.
Retrospectives are autopsies
Ninety minutes on what broke, four minutes on what worked. The team leaves with a precise map of its failures and a vague sense of its strengths — then wonders why morale tracks downward in a year that went well.
Doing nothing is scored as zero
Acting has a visible failure mode and waiting has an invisible one, so the ledger only ever books one of them. See The Golden Window for what the invisible column actually costs.
None of this argues for optimism as a posture. It argues for symmetry: if you would demand evidence for "this will work," demand the same for "this will fail," and put both columns on the same page. The neighbouring tools that do the arithmetic properly are The Expected-Value Tree and The Base Rate Machine; for why the pessimistic read feels more rigorous than it is, see The Survivorship Bias Lens.
Go deeper
Optimism is not a mood. It is a discipline about time.
The honest version of digital optimism is not "it will be fine." It is a claim about sequencing: the costs of any real change are front-loaded, specific and felt by named people, while the benefits are back-loaded, diffuse and compounding. A room that understands negativity bias can hold that shape without either denying the costs or being governed by them — which is the whole job. The S Curve shows the same asymmetry as geometry, and The Cost of Waiting prices the column nobody keeps.
Talk to us about running this session