Murray State University · Part two · Where it is going

AI Winters: Boom and Bust

Murray State University
Demo 4 of 17 · Part two · Stage 1

Read the claim. Place it in time.

Is this a forecast you can grade, or an assessment you can only judge by its own moment?

Before you start. A forecast names an outcome and a deadline; an assessment or untimed ambition describes conditions at a moment with no scoreable future window. Ten sourced claims follow, shown with the speaker and date hidden — five forecasts, four assessments or untimed ambitions, and one warning.

Predict

Guess the year this claim was made, then decide: is it a forecast you can score, or an assessment you can only classify?

Try

Drag the year slider, choose a verdict, then press Lock it in to reveal the year, speaker and source.

Takeaway

A forecast needs an outcome and a deadline; an assessment needs evidence about its own moment — grading one as the other makes history look more certain than it is.

Ten claims, 1950–2024

Check yourself

Two of the ten cards are still unresolved ("open"). Should an open forecast be marked wrong?

Demo 4 of 17 · Part two · Stage 2

Two winters. Different pathways.

Do the two AI winters share one fixed cause-and-effect sequence, or different pathways that happen to rhyme?

Before you start. Stage 1 asked you to date individual claims. Each event on this timeline carries a lane — claims and ambitions, institutional bets, constraints, reassessment, or withdrawal — plus landmark results shown separately as ground truth. The two shaded bands mark the conventional, soft-edged date ranges historians use for each winter.

Predict

Scrub from 1950. In each winter, does an institutional bet or a named constraint fire first?

Try

Drag the date slider (or step with ‘ and ’) and select a dot for its full text and source.

Takeaway

The five lanes organize evidence for comparison; they do not prove that every AI contraction follows one fixed order.

Explore 1950–2026Drag the date or step between events
Live event and withdrawal counts, computed here as you scrub
2026showing up to
0events so far
0withdrawal events

Read the six timeline rows Sourced every dot's date, text and lane resolve to a cited source
Why the winter dates have soft edges
Four funding examples with different accounting bases

There is no continuous, sourceable series of AI funding across these seventy-six years — particularly not for US spending before 1980 — so this demo does not draw one. These are four sourceable funding examples with different bases: programme budgets, programme expenditure, and a private funding round. They use three currencies, different price years, and no common accounting basis. Read them as dated evidence, not as a combined total or trend.

Check yourself

In both winters, a named institutional bet appears before a withdrawal on the timeline. Does that mean big funding always precedes an AI contraction?

Demo 4 of 17 · Part two · Stage 3

What is similar? What is different?

Which of the five comparison lenses still shows a real similarity once you separate description from prediction?

Before you start. Choose one lens at a time — claims, money, constraints, reassessment, or withdrawal — then read the first winter, second winter and today side by side. The Today column is deliberately left open where the evidence does not yet settle it.

Predict

Pick the "Constraints" lens. Does "Today" show an established limit, or an open question?

Try

Select a lens button, then press Details on any era cell for its full explanation and sources.

Takeaway

An open cell is evidence only that a limit or a withdrawal has not (yet) been established — not a prediction that one is coming.

Sourced every era's headline text resolves to a cited source in its Details

Check yourself

The "Today" column leaves two cells open. Does that mean an AI winter is currently impossible?

Demo 4 of 17 · Part two · Stage 4

What can history tell us about today?

Which rhyme still holds once you weigh it against a genuine difference?

Before you start. A rhyme is a pattern that recurs across the two winters and now; a difference is a way today is not like either winter. Expand a card in each column before deciding whether an analogy survives both.

Predict

Pick one rhyme and one difference. Does the rhyme still hold once you weigh the difference against it?

Try

Expand a card in each column, then open Every source to check where each claim comes from.

Takeaway

A historical rhyme is a reason to ask a question, not a reason to expect the same ending — each claim about today needs its own evidence.

Sourced every rhyme and difference resolves to a named source below

Check yourself

Today's AI systems have far more users than either historical boom. Does that mean this one is not heading for a contraction?

AI Winters: Boom and Bust · Bryant Harrison · Murray State University
One HTML file · runs offline · source-backed historical inquiry
Printable guide

Guidehow to use this demo

Has this happened before? AI has contracted twice. The histories rhyme, but neither is a script for the next one.

Where this sits: Demo 3 measured a benchmark under stated conditions; this demo asks you to read a historical claim the same way — as a forecast with a deadline, or an assessment of a moment — before judging it.

  1. Claims. Guess the year and verdict for ten sourced claims, then press Lock it in to reveal the answer and its source.
  2. History. Drag the timeline slider or step with ‘ and ’ to inspect landmark results and five comparison lenses across 1950–2026.
  3. Compare. Choose a lens, then open each era's Details for the full explanation and sources.
  4. Today. Expand a similarity and a difference, then decide what evidence supports your own analogy.

Getting around

The four tabs switch stages. Details opens the assumptions and sources for the current stage. Guide reopens this card; Settings holds Open Presenter Notes, Presentation mode, Reset, and Run the self-test.

Time

Fourteen to nineteen minutes for all four stages; six minutes for the short version described in Presenter Notes.

Settings

Presenter notes

The printable session plan, on screen: what to say at each stage, which predictions to keep if you are short of time, the questions you will get, and a six-minute cut. This is the printable guide itself, not a summary of it.

Presentation mode

Scales the whole page up one notch for projecting to a room, and puts a Notes button in the header.

Reset

Back to the start: every answer cleared, stage 1, the timeline rewound to the full span, every opened card closed, the self-test output cleared. The page does not reload. Presentation mode stays as you set it.

Proof

Checks that every claim on screen resolves to a declared source, that no source is cited and never used, that every source URL points at a specific page rather than a bare host, that the timeline runs forward in time, that each winter contains at least one funding withdrawal, that the counts quoted on the scorecard match the cards actually in the deck, and that the scoring refuses to mark an unresolved prediction. Results appear here and in the browser console.

DetailsClaims

What is live, and what is sourced

The scoring is computed here; the claims are quoted from named sources

Live your miss distance and verdict correctness

The moment you press Lock it in, this page computes how many years your guess missed by and whether your verdict matches what happened, from the year and verdict you chose. Nothing is looked up for that comparison.

Sourced the ten claims themselves

Each quotation, its speaker, date and outcome are drawn from a named, checkable source and marked primary (checked against the document itself) or reported (a specific secondary source that names its own underlying scholarship) — for example Turing's 1950 paper (primary) versus Minsky's 1967 line, routed here through Quote Investigator's page references (reported).

How it works

A claim's kind (promise, assessment, or warning) decides whether it can be scored at all: only a time-bounded promise or warning receives an outcome verdict. An assessment or untimed ambition is classified, not graded against a later result, because it never named a deadline to miss.

Assumptions and simplifications

  • The kind (promise / assessment / warning) is this activity's own classification, made from what each quotation actually says, not a category the original speaker used.
  • This is a curated teaching deck of ten claims, not a representative accuracy sample of optimists, pessimists, or the field.
  • Two forecasts are left deliberately open (unresolved as of this build) and are never scored as failures.

Sources

Boundary: this stage does not claim these ten claims are a representative sample of forecasting accuracy in AI or any other field. It is a curated teaching selection chosen to make one distinction — forecast versus assessment — visible.

Details

Presenter Notesthe printable guide, on screen

Demo 4 of 17 · Part two · 14–19 min full, 6 min cut

AI Winters: Boom and Bust — presenter guide

One HTML file, offline, no AI at runtime. You need: a screen to project or share; no other materials.

Learning goals

Not stated objectives — what the activity itself demonstrably shows, one per stage:

  1. A forecast needs an outcome and a deadline; an assessment needs evidence about its own moment — grading one as the other makes history look more certain than it is.
  2. The five lanes organize evidence for comparison; they do not prove that every AI contraction follows one fixed order.
  3. An open cell is evidence only that a limit or a withdrawal has not (yet) been established — not a prediction that one is coming.
  4. A historical rhyme is a reason to ask a question, not a reason to expect the same ending — each claim about today needs its own evidence.

Where this sits on the path

Demo 3 leaves the room reading a benchmark result under its own stated conditions rather than extrapolating from two points. Fast progress like that is exactly what raises the question this demo answers: is this a bubble? History answers with evidence, not a script — and this demo hands forward the habit of classifying a claim, forecast or assessment, before judging it.

One question for the session

Has this happened before? AI has contracted twice, through different combinations of technical limits, institutional choices and markets. The histories rhyme; neither is a script for the present.

The one sentence to land: a prediction needs a deadline and an outcome; an assessment needs evidence about its own moment. Do not grade a historical claim until you know which kind it is.

The shape of the session

StageWhat happensTime
1 — ClaimsFive time-bounded forecasts, four assessments or untimed ambitions, and one warning, with the speaker and date hidden. Score forecasts; classify the unscored context.7–9 min
2 — HistoryScrub the timeline. Five lenses compare historical pathways without forcing a sequence.3–4 min
3 — CompareCompare the two winters and the present under the same lenses. Two current questions stay open.2–3 min
4 — TodayRhymes against differences. Close on the checklist.2–3 min

Stage 1 — Claims

Question. Is this a forecast you can grade, or an assessment you can only judge by its own moment?

Predict prompt. Guess the year, then say whether the claim is a scoreable forecast or an assessment to classify — before Lock it in reveals anything.

Activity. Do the first two cards yourself, out loud, so the room sees the mechanic. Then hand it over — ask for a show of hands on the year before you drag the slider, and a verbal verdict or classification before you press Lock it in. For the four context cards, choose Assessment or untimed ambition when the claim supplies no scoreable future window. The room being wrong together is the experience; a silent room reading a screen is not.

Expected observation. Three of the five time-bounded promises missed outright, one arrived thirty years late, and the deliberately unscored ones (the 1958 press claim, ALPAC, Perceptrons, Lighthill, and Amodei 2024) pull the room toward wanting to grade them anyway — that pull is the misconception to name.

Check-yourself answer key. "Should an open forecast be marked wrong?" — No: an unresolved forecast stays unscored rather than counted as a failure.

Teaching limit. This is a curated ten-claim deck, not a representative accuracy sample of optimists, pessimists, or the field.

The deck, and what each card is for

YearCardTypeWhat it is for
1950Turing — fifty years to the imitation gameopen forecastIts about-2000 bar was missed; the current status remains contested. Use it to separate a past deadline from an unresolved present judgment.
1958“walk, talk, see, write, reproduce itself”untimed ambitionAn expansive press claim without a deadline. Classify it; do not pretend it supplies a scoreable outcome window.
1957Simon & Newell — chess champion in ten yearspromiseThe most useful card in the deck. Right destination, wrong by a factor of four. Pause here.
1960Simon — any work a man can do, in twenty yearspromiseTwenty years lands in 1980, the bottom of the first winter. Also: it is usually misdated to 1965.
1966ALPAC — machine translation is not nearassessmentDistinguish a near-term economic assessment from a forecast. The report also recommended continued computational-linguistics research.
1967Minsky — substantially solved within a generationpromiseA reported quotation attributed to his textbook. Use it to explain why the app labels this source route reported.
1969Perceptrons — neural nets read as a dead endassessmentThe proofs were right; the history of how the field received them is disputed.
1973Lighthill — no major impact so farassessmentA retrospective judgment on impact and constraints, not a permanent forecast.
1984Schank & Minsky — a winter is comingwarningThe card that saves the demo from being anti-hype. They were right, three years early.
2024Amodei — “could come as early as 2026”openDeliberately unscored. The year named is the year you are standing in.

Short of time? Keep four: Simon & Newell 1957 (over-promise, right destination), ALPAC 1966 (assessment versus forecast), Schank & Minsky 1984 (a prediction that was right), and Amodei 2024 (still open). Skip to Stage 3 and close.

Stage 2 — History

Question. Do the two AI winters share one fixed cause-and-effect sequence, or different pathways that happen to rhyme?

Predict prompt. Scrubbing from 1950, does an institutional bet or a named constraint fire first in each winter?

Activity. Drag the slider slowly, or step with ‘ and ’ to stop and read one event. Tap any dot for its full text and source. Use the rows as comparison lenses — claims, institutional bets, constraints, reassessment, withdrawal — and ask which causes matter in each episode and which do not carry across. Read the current entries as evidence about today, not a third historical run.

Expected observation. The order is not identical between the two winters: the first winter's reassessment (ALPAC, then Lighthill) came from commissioned reports years apart, while the second winter's reassessment was the market itself — the LISP hardware collapse, with no equivalent report. Two histories, not one template.

Check-yourself answer key. "Does a named bet before a withdrawal in both winters prove a rule?" — No: two episodes sharing an order is not a proven causal sequence.

Teaching limit. No continuous, sourceable AI funding series exists across these years; four discrete funding examples with different currencies and accounting bases are shown instead of one invented total.

Do not let the lenses turn into a prophecy. The correct line is: “these comparisons tell us what to investigate. They do not establish that a withdrawal will occur.”

Stage 3 — Compare

Question. Which of the five comparison lenses still shows a real similarity once you separate description from prediction?

Predict prompt. On the "Constraints" lens, is "Today" an established limit, or an open question?

Activity. Choose one of the five lens buttons; its navy fill marks the selection. Compare the three era headings, then open each era’s Details for the full explanation and named sources. Escape closes the panel and returns focus to its button. Compare causes, not step numbers.

Expected observation. Two cells in the "Today" column stay open — no single binding technical constraint is established, and no comparable broad contraction is established either — while the claims, money and reassessment lenses do resolve to sourced present-day evidence.

Check-yourself answer key. "Does an open cell mean a winter is currently impossible?" — No: open means not yet established either way, not ruled out.

Teaching limit. Calling a present-day cell "open" is this page's own reading of the evidence available at build time, not a citation that says "unresolved."

“Hindsight makes old constraints easier to see. It does not prove that today has the same constraint or that it must end the same way. Ask what evidence would support the claim.”

Stage 4 — Today

Question. Which rhyme still holds once you weigh it against a genuine difference?

Predict prompt. Pick one rhyme and one difference — does the rhyme survive once you weigh the difference against it?

Activity. Expand a similarity and a difference, giving both columns roughly equal time; the supporting statements start closed so the room investigates one claim at a time. Close on the checklist, not a verdict: is it a forecast or an assessment, what evidence supports it, and what evidence would move it.

Expected observation. The strongest rhyme (an evaluation warning about SWE-bench Verified) and the strongest difference (measured scaling behaviour and reported consumer reach) do not cancel each other — each still needs its own evidence, which is the point.

Check-yourself answer key. "Does a bigger user base rule out a contraction?" — No: user count answers a different question than whether funding, evaluation trust, or returns hold up.

Teaching limit. Historical comparison is not a prediction; this curated set of near-term forecasts that missed does not establish a representative error rate for optimists, pessimists, or AI research generally.

Misconceptions

"An unresolved forecast should count as a failure." No. Only a forecast whose window has closed one way or the other gets a verdict; an open claim stays unscored (Stage 1).
"The same order appearing in both winters proves a fixed sequence." No. Two historical cases sharing a pattern is a reason to look closer, not a general law about every contraction (Stage 2).
"An open cell in the 'Today' column means a winter cannot happen." No. Open means the evidence has not established a limit or withdrawal yet — not that one is ruled out (Stage 3).
"More users than any earlier boom means this one avoids a contraction." No. Scale of adoption is a genuine, sourced difference, but it answers a different question than whether funding, evaluation trust, or returns hold up (Stage 4).

Discussion questions

  • If someone states a confident AI timeline today, what would you need to see before you could score it — and by when could you score it?
  • The first winter's reassessment came from a commissioned report; the second came from a market collapse with no report at all. Does the absence of a report today mean nothing is wrong, or only that the evidence would look different if something were?
  • Pick one "Today" claim from Stage 4. What single piece of evidence would move your confidence in it the most, in either direction?

Questions you will get

“So is it a bubble or not?”

Refuse the binary, in one sentence, then give the useful answer: current funding, reported usage, technical reliability, and returns are distinct questions. Earlier eras also had deployed systems and organisational users, though current consumer reach is much larger. A financial correction and a technological dead end are different claims and need different evidence.

“Aren’t the winter dates made up?”

Partly, and the demo says so. The bands are conventions for broader contractions, not dates stamped by a single event. The 1987 hardware collapse is a useful second-winter anchor; the first has more distributed causes and withdrawals, including but not limited to ALPAC. The ends are also conventional and historians disagree; that is why the bands have soft edges and explanatory notes.

“Where did you get this?”

Open Settings → Run the self-test in front of them. It re-derives, live, that every claim resolves to a declared source, that no source is cited and never used, and that the counts quoted on screen match the data. Then open Stage 4, which lists every source under Every source and, separately, everything researched and cut for lack of one. Both supporting sections open on request.

“Why is that quote from a quotation-checking website?”

Because the Simon and Minsky lines are routed through a quotation-checking source that cites printed books and page references. This build has not inspected those original pages, so it labels the claims reported rather than primary. That distinction is part of the lesson: a plausible citation route is not the same as direct verification.

Presenter controls

  • Turn on Settings → Presentation mode for projector-sized type.
  • Run the self-test once so you know it is green before anyone asks.
  • Between rooms, Settings → Reset puts the demo back to its start without reloading the page — answers cleared, Stage 1, timeline rewound, opened cards closed — and keeps Presentation mode on so the projector does not change under you.
  • The app saves nothing anywhere, so an ordinary reload resets it too.
  • No network is needed at any point. If the room’s wifi is down, nothing changes.

The buttons, top right

  • ? — Guide: the short how-to panel that opens on first load, reopenable at any point. (Not this document.)
  • ⚙ — Settings: Open Presenter Notes (this document, on screen), Presentation mode, Reset, and Run the self-test.
AI Winters: Boom and Bust — Bryant Harrison, Murray State University. Sources checked September 2026. Every figure on screen links to its origin; everything that could not be sourced is listed in Stage 4 under “What was cut, and why”.

Open the printable copy (presenter-guide.html) — the same document, laid out for print; the PDF beside it is a render of that file.
Bryant Harrison · Murray State University