Where is AI going?
A weighted read of 58 open questions — a distribution of expert views — not a forecast.
The AI Speedometer: how experts think the next few years will play out — a distribution, not a forecast.
Experts hold a steady pace — and no world has run away.
A · Read
The scenario read
—
B · Read
Pace
—
All 58 units · panel placement
Tap to open
01 · Placement matrix
Eighty experts, fifty-eight questions.
Every read above is a pure function of this one matrix. The surface stays blurred to a density read; per-cell provenance and named placements live in the drawers.02 · Methodology
How the read is built.
How roughly 80 experts’ public positions become seven scenario reads, and why none is crowned the winner.- What deslop.media measures
- Each expert is placed on five constraints, not on a scenario: whether AI capability keeps improving; whether the build-out of power, chips, and financing keeps scaling; whether the spending turns into real revenue; whether the job market absorbs the disruption; and whether government stays out of the way. For each constraint, deslop.media reads one headline question plus the handful of specific sub-questions beneath it, then blends them — half the headline, half the average of the specifics — into a single stance: holds, denies, or mixed. Every stance rests on a dated public quote. Where an expert hasn’t spoken to a constraint, that cell stays empty, and an empty cell is never filled with a guess.
- How positions become scenarios
- A scenario is a recipe: one specific setting for each of the five constraints. “Capability holds, build-out holds, revenue denies” is the Monetization Gap. The tempting shortcut is to take each constraint’s odds and multiply them together — a shortcut deslop.media refuses, because multiplying conjures combinations no expert actually holds and puts weight on futures nobody believes in. Instead each expert’s real five-constraint row travels intact and is matched against every recipe. Contradict the recipe and the expert routes nowhere; fit it and the expert is credited in proportion to how much of the recipe they affirmed; stay silent and they neither help nor hurt. Each scenario’s score is then divided by the experts who engaged its constraints at all, so a scenario isn’t penalized for thin coverage.
- Why no scenario wins
- On today’s evidence the leading scenarios sit inside one another’s margin of error. The read is contested, so nothing is crowned. What the bars do show is solid: how consistently the panel routes to each scenario, which cleanly separates the cluster where the engine keeps running from the cluster where it stalls. That separation is the signal. The ranking within each cluster is noise, and the figure doesn’t pretend otherwise.
- Pace
- Ordering the seven scenarios coldest (the frontier stalls) to hottest (fastest, most disruptive) and taking their weighted centre of mass yields one needle on a scale from slow to fast. Today it reads steady. “Cold” means the frontier stops advancing while today’s tools keep spreading; it never means AI runs backward.
- Where the read is thin
- Coverage is uneven. The capability and commercial constraints are well answered; the politics constraint is answered by only a couple of experts, so any political signal is held lightly and flagged rather than inflated into a panel verdict. Experts whose positions match no single recipe are shown as unmatched, not hidden. The honest gaps are part of the figure.
03 · Sources
The citation feed.
A live, append-only stream of named public statements — and the placement deslop.media reads each as. Each card carries the question’s 12–24 month resolution test. The great majority of matrix cells are silent, and silence contributes no card; this feed draws only on the sourced, non-silent placements. Click a matrix cell to jump in.
Live · append-only
·
0 sourced placements
·
0 holds
0 denies
·
silent cells carry no card — absence is the signal
·
last appended —
Read honesty Every card is public-corpus inference — a verbatim, dated, public statement that deslop.media reads onto a question. None are statements the person gave deslop.media. The stance verb attaches to a question’s resolution test, never to a scenario or camp; a quote is as often a deny anchor as a hold.
↘ From the placement matrix
Each active matrix cell deep-links [L#] to a card. Click one — it force-loads the card even if lazy-held, scrolls it in, and flashes it. Silent cells dead-end nowhere by design.
Group
Stance
Within group
04 · Questions
How readers ask about this.
Plain answers to the questions this figure tends to raise — for the reader, and for the agent reading on their behalf.Is this a prediction?
No — it’s a map of disagreement, not a forecast. It shows where invited experts land across the open questions. On current evidence the read is contested: the panel leans toward the engine holding but is genuinely divided on which scenario follows, so the figure crowns none of them. As the evidence shifts, the weight shifts with it.
Why don’t the experts have a view on most questions?
Because honest uncertainty is the point. On any given question, most experts haven’t taken a position we can stand behind — so we record the silence rather than invent an answer. Where that silence is wide, the question is genuinely open, and that is itself a finding.
Who placed these — and can I?
deslop.media has synthesized the perspectives of 80 experts across AI — scientists, technologists, business leaders, economists, pollsters and others. Two further views are planned: your own, by answering the questions yourself, and a deslop.media forecasting model. The source control sits at the top of the figure.
Why are there no numbers on the chart?
Because the honest signal is the shape of the disagreement, not a false-precise figure. Weight and lean read through size and position; the underlying counts and the named placements are a tap away on any element.
How often does this update?
deslop.media is committed to keeping the content fresh. If you see something missing, email [email protected].
Can an agent read this?
Yes — the same analysis is published in a structured, machine-readable form alongside the page, so an agent can query the distribution directly.
05 · Machine channel
The same read, typed for agents.
deslop.media is read by agents first — but a human rarely reads JSON, so it sits last. This is the figure’s quantitative mirror, served as data.json beside the chart and never printed on it.data.json · scenario-pyramidagent ready