Sputnik 1

Filed as moment, and this is the least contested filing in the canon — which is exactly why the entry has to spend its length somewhere else. build.py states the kind as "Something that happened and changed what people expected next." llama-weights-2023 fixed the operational test as a date on which something visibly happened in public, and arpanet-1969 had to argue the words "in public" for four paragraphs. This entry needs none of that. There is a date to the second (4 October 1957, 19:28:34 UTC), a place (Site No. 1/5 at the 5th Tyuratam range, later Baikonur), an object with a mass (83.6 kg), and a public so wide that amateurs with shortwave sets heard the transmitter themselves on 20.005 and 40.002 MHz for twenty-two days. printing-press-1440 failed the moment test because it could produce no day at all. This one can produce a minute.

The difficulty is the opposite one, and it is the reason this file is long. What a 2026 reading actually meets is almost never the satellite. It is the phrase "Sputnik moment", used as a premise — a compressed argument that a foreign technical demonstration has revealed a gap, and that the correct response is a large public buildout. That phrase is an interpretation living inside a moment, and it is wrong about the moment it is named after in at least three specific ways. The house rule for limit entries — carry what the thing actually says next to what it is repeatedly made to say — is not required here, but it is the only way this entry earns admission, so section 4 does it anyway.

---

On descends_from, and the edge this entry cannot draw

descends_from is empty and honestly so. Nothing in canon/ precedes Sputnik causally. dartmouth-1956 is fourteen months earlier and is not an ancestor of it; von-neumann-singularity-1958 is the next year and shares only a decade. The satellite came out of Soviet ballistic-missile work, which this canon does not hold and which this entry does not propose adding.

The edges that matter run downward, and one of them I am forbidden from writing. arpanet-1969 closes with this: "The entry I most want an edge to and cannot have is sputnik-1957, and when that entry is written this one should be revisited, because the correction above belongs on the edge between them rather than inside either file." That entry is now written. arpanet-1969.md still carries descends_from: [], and jobs/canon.md forbids this job from touching another entry's file, so the edge remains undrawn and this paragraph is the only record that it should exist. A session or job that is permitted to edit arpanet-1969.md should add sputnik-1957 to its descends_from and delete its closing paragraph. darpa-grand-challenge-2005 has the same claim on the same grounds, two agency-generations further down.

I have not invented any id. Everything named above already exists in canon/.

---

What it is

The object

Sputnik 1 was an aluminium sphere 58 cm across and 83.6 kg in mass, with four whip antennas and, inside, three silver-zinc batteries, a thermal switch, a fan, and two radio transmitters. It carried no instrument. Its scientific yield was indirect — the way the signal changed as it crossed the ionosphere, and the way its orbit decayed, which said something about atmospheric density at altitude. The beep was not a message. It was a carrier wave whose pulse length shifted with the internal temperature and pressure, so the sound itself was the telemetry.

It orbited every 96.2 minutes. The batteries died on 26 October 1957, twenty-two days in. The sphere itself reentered on 4 January 1958 after 1,440 orbits. What most people actually saw in the sky was not the satellite, which was small and dim, but the spent core stage of the R-7 tumbling in the same orbit, which was very much larger and easily visible. The most-watched object of the Space Age's first month was a piece of discarded rocket.

The launcher is the part that mattered. The R-7 was not a space vehicle adapted to war; it was the world's first intercontinental ballistic missile, adapted to carry a beeping sphere because Korolev pushed for it and because the International Geophysical Year gave a civil pretext. The satellite was a by-product of a weapons programme, and everyone in a position to know understood that at once.

What Eisenhower actually thought, which is not what the country thought

The United States was not surprised in the way its newspapers were. Eisenhower had been reading U-2 photography and telemetry intercepts of R-7 tests for months; the Soviets had flown the missile successfully in August 1957 and had said so. More important, the administration had a reason to want the Soviets to go first.

NSC 5520, the draft statement of policy on the US scientific satellite programme dated 20 May 1955, gives it plainly: a small scientific satellite would establish the international legal precedent of "freedom of space" — that a body in orbit may pass over a country without violating its airspace, the way an aircraft may not. Eisenhower had already tried to get this conceded directly, proposing mutual aerial inspection at the Geneva summit on 21 July 1955; the Soviets rejected it as an American attempt to accumulate target information, which it partly was. The civil satellite programme (Vanguard) was the second attempt at the same objective by other means, running deliberately separate from the classified reconnaissance-satellite work that became CORONA.

So when the Soviets launched first and no one protested the overflight, the precedent was established by the party that would have been the one to object. Donald Quarles told the President on 8 October 1957 that the Russians had "done us a good turn, unintentionally, in establishing the concept of freedom of international space." At his press conference the next day, 9 October 1957, Eisenhower said that "from what [the Soviets] say, they have put one small ball in the air" and that it represented "no additional threat to the United States."

He was, on the narrow question, right. He also completely failed to sell it, and the gap between the President's assessment and the public's is the actual historical event.

The panic, and the two things that fed it

Two things converted surprise into crisis, and neither is the beeping sphere.

Sputnik 2, on 3 November 1957, weighed 508.3 kg and carried a dog. The mass is the whole point: a booster that can put half a tonne into orbit can put a thermonuclear warhead onto a continent. Sputnik 1 was a novelty; Sputnik 2 was a throw-weight demonstration, one month later. The satellite that gave the analogy its name is not the one that carried the military information. (Laika died within hours when thermal control failed. The USSR maintained for forty-five years that she had died painlessly after several days; a scientist involved admitted the truth in 2002.)

Vanguard TV-3, on 6 December 1957, rose about four feet off the pad at Cape Canaveral, lost thrust and exploded, on television. Its payload was 1.36 kg and 15 cm across. The newspapers called it Flopnik, Kaputnik and Stayputnik. The American answer arrived on 31 January 1958 — Explorer 1, on a Juno I, whose single instrument discovered the Van Allen radiation belts, a better piece of science than anything the Sputniks returned. By then the political fact was fixed and the science did not move it.

What the panic bought, with dates and figures

Within eleven months of the launch the United States created two agencies and passed the largest federal intervention in education in its history to that point.

• ARPA, by DoD Directive 5105.15, issued 7 February 1958 by Secretary of Defense Neil McElroy — four months and three days after the launch. First director Roy Johnson, out of General Electric. Its first three research thrusts were space technology, ballistic missile defence and solid propellants.
• NASA, by the National Aeronautics and Space Act signed 29 July 1958; the agency began operations 1 October 1958, absorbing the NACA whole.
• The National Science Foundation's appropriation. By NSF's own history: "In fiscal year 1958, the year before Sputnik, the Foundation's appropriation had leveled at $40 million. In fiscal 1959, it more than tripled at $134 million." The single largest beneficiary was education — $62,070,000 for FY1959, more than the Foundation's entire budget the year before.
• The National Defense Education Act, P.L. 85-864, signed 2 September 1958: over $1 billion, ten titles, student loans, graduate fellowships, laboratory equipment, guidance counsellors, foreign-language and area studies, and — Title VII — the first federal money for educational technology research.

That is a real institutional response at speed, and it is the reason the analogy exists at all. Two details are almost never carried with it.

The first: the NDEA was not written after Sputnik. The bill and its constituency long predated the launch; what Sputnik supplied was a name and a majority. The Senate committee clerk Stewart McClure suggested selling an education bill as a defence measure, and Lister Hill and Carl Elliott renamed it accordingly. The crisis did not produce the policy. It opened a window that people who had been holding the policy for years walked through.

The second: the money came with a loyalty test. NDEA loan recipients had to swear an oath of allegiance and sign an affidavit disclaiming belief in overthrowing the government. Universities fought it for four years; the disclaimer affidavit was repealed in October 1962 under Kennedy, the oath retained. A national mobilisation against a foreign technical rival arrived with a demand that individual students account for their beliefs. Anyone reaching for this precedent should reach for that part too.

The line from here to AI, which is real and is not straight

ARPA was created to run America's space programme and lost that mission almost immediately — the transfers to NASA through 1958–60 are set out in arpanet-1969 and are not restated here. What is relevant is the residue: an agency a few years old, stripped of the portfolio the panic created it for, looking for work. It found computing. The Information Processing Techniques Office was established in 1962, with J. C. R. Licklider as its first director from October of that year. Project MAC launched on 1 July 1963 on a $2 million ARPA grant. The National Research Council's Funding a Revolution states the outcome flatly: "DARPA provided the bulk of the nation's support for AI research" from the 1960s through the 1990s.

Two qualifications keep this from becoming a fairy tale.

AI did not need Sputnik and did not wait for it. The Dartmouth workshop (dartmouth-1956) ran in the summer of 1956, fourteen months before the launch; logic-theorist-1956 was demonstrated that same summer; shannon-chess-1950 and turing-1950 are older still. The field existed. What Sputnik bought was not the idea but the funding regime — a decade in which a programme manager could back an approach without justifying it against a military requirement.

And that regime was closed by law while the field was still young. The Mansfield Amendment (§ 203 of P.L. 91-121, November 1969, carried into the Defense Procurement Authorization Act of 1970) required DoD to support only basic research with "a direct and apparent relationship to a specific military function or operation." The blank cheque the panic wrote had a term of about eleven years. What came after it — the accounting, the deliverables, the disappointment — is lighthill-1973 and expert-systems-collapse-1987.

---

Why a reading would cite it

The proposal line in canon/proposals.md reads: "sputnik-1957 — Sputnik, and ARPA founded four months later. Cite when: national rivalry is claimed to accelerate a field — this is the case where it demonstrably did, and the agency it created went on to fund most of early AI and then ARPANET. The same entry is the warning: the acceleration was real and the panic that produced it was mostly wrong about the specific threat. proposed 2026-08-17."

That is right, and it survives contact with the sources, with one adjustment argued in section 4: demonstrably did is true of the money and much weaker for the mechanism.

The citation occasion is live in the most literal sense. The reading of 19 August 2026 carries, as its own lead, two governments each putting a hole in the same wall in the same week — ByteDance and Tencent each taking roughly 10,000 Nvidia H200s with US clearance reportedly extending to 100,000 apiece, while Beijing tells the same firms to keep the chips outside the mainland to protect domestic chipmakers. Every item in the Concentration lens for months has been an instance of the same claim: that national rivalry is the engine, and that the correct read of any capability event is which country gained. This entry is the base case for that claim, with the receipts and the failure modes attached.

There is a second occasion, narrower and more frequent: whenever anyone says "Sputnik moment." They have said it about AI at least twice at scale — Kai-Fu Lee about AlphaGo, Marc Andreessen about DeepSeek — and each time the phrase did argumentative work that its speaker did not have to defend. A reading that meets the phrase can either let it pass or make it pay, and section 4 is what it costs.

What this entry is not for. It is not evidence, it does not move the needle, and it has no bearing on whether a given week's chip export decision was good or bad. It is a reference class: the best-documented case of a technological shock producing a state response, held next to what the shock actually was.

---

What it got right, and what it got wrong

Not required for moment, but this entry is half prediction in effect: every invocation of it forecasts a response. So the claims get graded — the claim in its own words, when it was made, when it came due, and what happened.

The 1957–58 threat assessment

Claim (NIE 11-10-57, December 1957): the Soviets would "probably have a first operational capability with up to 10 prototype ICBMs" at "some time during the period from mid-1958 to mid-1959." Claim (August 1958): an operational capability with 100 ICBMs some time in 1960, and perhaps 500 ICBMs "some time in 1961, or at the latest in 1962." Due: 1961–62.

What happened: In January 1961 McNamara and Gilpatric examined CORONA satellite photography. R-7 launchers are enormous and would have been unmistakable; none appeared. In February 1961 McNamara stated there was no evidence of a large-scale Soviet ICBM effort. A September 1961 estimate put the Soviet force at no more than 25 ICBMs, on 10–25 launchers. The R-7 itself was a poor weapon: roughly twenty hours to prepare for launch, unable to hold alert beyond a day because of its cryogenic oxidiser, with about six deployed at peak across five sites, and about thirty ever built for the ICBM role.

Grade: wrong by one to two orders of magnitude, and wrong in the direction of alarm. The launch was real, the capability it advertised was real, and the force behind it was almost entirely absent. The public inference — they are ahead of us — was not merely unproven; the reverse was true and the US government had photographs proving it within three years. This is the single most important thing in the file. A dramatic public demonstration is evidence about a demonstrated capability and very weak evidence about a fielded one.

"China's Sputnik moment" — Kai-Fu Lee, 2017/2018

Claim: AlphaGo's defeat of Ke Jie in May 2017, watched by a reported 280 million people in China, was China's Sputnik moment, and would produce a state-led national response. Made 2017–18, in AI Superpowers. Due: the State Council's own New Generation Artificial Intelligence Development Plan (July 2017) set dated benchmarks at 2020, 2025 and 2030.

What happened: the plan is real, dated, and was followed by a decade of industrial policy. By 2026 the outcome is visible in the readings themselves — Chinese open-weights models at or near frontier capability, the largest agent-mediated commerce surface anyone has stood up, a domestic humanoid-robot maker listing publicly at a valuation far ahead of its shipments.

Grade: the strongest instance of the analogy holding, with a caveat that matters. The causal arrow is not established. The plan appeared roughly two months after the Ke Jie matches — faster than any policy of that scope can be drafted — which is the NDEA pattern exactly: a programme already in motion, handed a name and a mandate by a public shock. The shock is much better evidence about timing and framing than about existence.

"AI's Sputnik moment" — Marc Andreessen, 27 January 2025

Claim: DeepSeek's R1, released 20 January 2025, was AI's Sputnik moment. Due: the analogy is not a claim about a model; it asserts a mechanism — foreign demonstration → national alarm → large public buildout. That mechanism had one to two years to show itself.

What happened, as of August 2026:

• The public research base contracted rather than expanded. The FY2026 request would have cut NSF by 56.9% (and non-defence R&D by roughly $42 billion, about 21%); Congress rejected most of that, passing something near a 5% cut to non-defence R&D and, per trade reporting, an NSF appropriation of $8.75 billion signed 23 January 2026. But appropriation is not obligation: NSF had made about 613 grants by April 2026, roughly 20% of the level at the same point in FY2021–2024. Compare the post-Sputnik number: $40m → $134m in a single year.
• The buildout that did happen was federal in name and small in money. The Genesis Mission executive order of 24 November 2025 directs DOE to build an American Science and Security Platform and explicitly invokes the Manhattan Project for scale. The appropriation attached to the relevant line (OBBBA § 50404) is $150 million through September 2026. ARPA's founding year is reported at around $520 million in 1958 dollars.
• The actual Sputnik-scale response was private. Hyperscaler capital expenditure for 2026 is estimated in the $700–800 billion range — an analyst aggregation, not a filing, and carried here as an order of magnitude only. Whatever else 2025–26 was, it was not a period of small investment.

Grade: the claim was right about the alarm and wrong about the mechanism, and the mechanism was the entire content of the analogy. Alex Tabarrok put the contrast on 11 July 2025: "The U.S. answered Sputnik with bold competition not an aggrieved whine that America had been ripped off and abused." A Stimson Center piece of 25 April 2026 titled Reverse Sputnik argues the same, that the response has been retreat where the precedent was expansion. The honest finding is sharper than either: the money did arrive, at a scale the 1958 Congress could not have imagined, and it arrived from private balance sheets answerable to shareholders rather than from a public agency answerable to a legislature. An analogy that promises a national mobilisation and delivers a capex cycle has not been half-confirmed. It has been answered in the negative on the only question it was asked — who ends up owning the capability — which is the axis this project is named for.

---

Commonly misused as

1. "We are behind, therefore we must spend"

What actually happened: the United States was probably ahead, and had chosen to look behind. Vanguard was deliberately slow and civil because Eisenhower wanted the freedom-of-space precedent set by a scientific satellite rather than a military one, and the missile work that mattered was classified. Brian Weeden, writing at CSIS on 17 December 2021, put the general case: "Calling something a 'Sputnik Moment' without understanding what really happened during the original Sputnik crisis is a bad idea" — the public and the media "drew the incorrect conclusion that the Soviets were much further ahead than they actually were." His corrected definition is deliberately awkward: the term fits when a country is actually ahead but concealing it, and the public overreacts anyway. Almost nobody who reaches for the phrase means that.

2. "A shock reliably produces an institutional response"

What actually happened: the shock produced a window, and organised constituencies who already had drafts walked through it. The NDEA was a pre-existing federal-aid-to-education bill renamed as defence. Mark Hlavacik, writing on 3 October 2023, made the general point from the record of declared Sputnik moments: some produce lasting change, others "are easily forgotten" — Obama's own 2011 invocation "did not result in a lasting public discussion" and its centrepiece proposal, an ARPA for education, never existed. The shock is not the mechanism. The prepared constituency is the mechanism. A reading watching a capability event should ask who already has legislation drafted, not how loud the alarm was.

3. "National rivalry accelerates the field"

What actually happened, in the case most often cited as proof: rivalry created an agency to run a space programme, that agency lost the space programme within two years, and the residue wandered into computing and funded most of early AI for a decade until Congress closed the arrangement by statute. The acceleration is real and the causal chain is not clean; arpanet-1969 documents that in detail and this entry defers to it. The defensible form of the claim is narrow: national rivalry reliably buys institutions, and institutions sometimes produce things unrelated to the fear that bought them.

4. The phrase's own inflation

Sputnik moments have been declared for education by Mitt Romney (May 2005: "Our generation has not had a Sputnik moment…yet. But our Sputnik is on its way"), by Janet Napolitano and Tim Pawlenty (April 2007), by Barack Obama in the State of the Union of 25 January 2011 ("This is our generation's Sputnik moment"), and for AI by Lee (2017) and Andreessen (2025); the 1983 A Nation at Risk report and the 2016 election have both been assigned the label retrospectively. A term applied this often to this many unlike events is a rhetorical instrument for converting an observation into a budget request. The correct handling in a reading is to name the speaker and the ask in the same sentence as the phrase.

5. And the smallest correction, which is the most quotable

Sputnik 1 was not a weapon, not a threat, and not the militarily informative object. It was 83.6 kg of aluminium with a radio in it and no instrument, whose visible companion in the sky was a piece of spent rocket. The satellite that told American analysts something real about Soviet capability was Sputnik 2, four weeks later, at 508 kg — and it is not the one the analogy is named after. Every "Sputnik moment" declared since has been named after the beeper.

---

Sources

Web research conducted 19 August 2026. Where a claim rests on a search-result summary rather than a document I read, it is marked.

Primary and institutional.

• National Science Foundation, "The National Science Foundation: A Brief History" (nsf.gov/about/history/narrative), read directly. Source for the verbatim "In fiscal year 1958, the year before Sputnik, the Foundation's appropriation had leveled at $40 million. In fiscal 1959, it more than tripled at $134 million", and for NSF's own framing that "Sputnik raised questions about the ability of the nation's education system to compete. Congress responded with the National Defense Education Act of 1958." The $62,070,000 FY1959 education figure is from search results over the same document set and was not read in the narrative itself.
• DARPA, "ARPA is born" (darpa.mil/about/innovation-timeline/arpa-is-born), read directly. Source for "On Feb. 7, 1958, Neil McElroy, the Department of Defense Secretary, issued DoD Directive 5105.15 establishing the Advanced Research Projects Agency (ARPA)" and for the three founding thrusts. Negative finding: DARPA's own page does not name Sputnik as the cause and does not state an initial appropriation. The Sputnik attribution, Roy Johnson's name and background, and the $520 million initial funding are all secondary (SpaceNews, 7 February 2018; encyclopaedic compilations) and should be read as such — the dollar figure in particular is repeated widely without a cited appropriations line.
• NASA History, "One Small Ball…" (nasa.gov/history/monograph10/), read directly. Source for Eisenhower's 9 October 1957 press-conference wording "from what [the Soviets] say, they have put one small ball in the air" and "no additional threat to the United States." Negative finding: the widely quoted continuation "does not raise my apprehensions, not one iota" is not in this text and I could not verify it against any primary; it is therefore not quoted above.
• Wikipedia, "Sputnik 1", read directly, for the launch timestamp (19:28:34 UTC, 4 October 1957), site, mass, diameter, transmitter frequencies, battery depletion on 26 October 1957, 96.20-minute period, 1,440 orbits and reentry on 4 January 1958, and for the U-2/telemetry forewarning and the freedom-of-space rationale. Encyclopaedic rather than primary, and the launch parameters are stable across every source checked.
• Wikipedia, "National Defense Education Act", read directly, for P.L. 85-864, the signing on 2 September 1958, the ten titles, the "over one billion dollars" total, the $183m (1959) → $222m (1960) annual figures, and the Title X disclaimer affidavit and its repeal in autumn 1962.

Secondary, and labelled as such.

• The NDEA's pre-Sputnik origins — that the bill preceded the launch, that Senate committee clerk Stewart McClure proposed reframing it as a defence measure, and that Lister Hill and Carl Elliott renamed it — is from search results across the US House "History, Art & Archives" record and Wayne J. Urban's More Than Science and Sputnik: The National Defense Education Act of 1958 (University of Alabama Press). history.house.gov returned HTTP 403 and was not read. Urban's book was not read; only descriptions of its argument.
• NSC 5520 (20 May 1955), the Geneva proposal of 21 July 1955, and Donald Quarles's 8 October 1957 remark on the Soviets having "done us a good turn, unintentionally, in establishing the concept of freedom of international space" are from search results over Roger Launius's writing, the Federation of American Scientists' space-management history, and Air University's The Air Force Role in Developing International Outer Space Law. The NSC document itself was not retrieved.
• The missile-gap estimates — NIE 11-10-57 of December 1957 ("up to 10 prototype ICBMs", mid-1958 to mid-1959), the August 1958 successor (100 ICBMs in 1960, "perhaps 500 ICBMs some time in 1961, or at the latest in 1962"), the McNamara/Gilpatric CORONA review of January 1961, McNamara's February 1961 statement, and the September 1961 estimate of no more than 25 ICBMs on 10–25 launchers — are from search results over the Arms Control Association (May 2011), Britannica and military-history compilations. No NIE was read in original.
• R-7 operational figures (about twenty hours to prepare, alert limited to a day by cryogenics, up to six deployed from late 1959 across five sites, ~30 built for the ICBM role) are from encyclopaedic compilations.
• Sputnik 2 (3 November 1957, 508.3 kg, Laika, thermal-control failure and the 2002 admission) and Vanguard TV-3 (6 December 1957, ~4 feet, 1.36 kg payload, "Flopnik"/"Kaputnik"/"Stayputnik") and Explorer 1 (31 January 1958, Juno I, Van Allen belts) are from NASA history pages and Astronomy reached via search summary rather than read in full.
• The AI-funding chain — IPTO established 1962 with Licklider from October, Project MAC on 1 July 1963 with a $2 million ARPA grant, "DARPA provided the bulk of the nation's support for AI research" (National Research Council, Funding a Revolution, 1999), and the Mansfield Amendment requirement of "a direct and apparent relationship to a specific military function or operation" — is carried from arpanet-1969, which sourced it to the NRC volume directly, plus search results confirming the Project MAC date and grant. ARPA's loss of its space mission is documented there and deliberately not re-derived here.
• Kai-Fu Lee's "China's Sputnik moment", the 280-million viewership figure, and the State Council's New Generation Artificial Intelligence Development Plan of July 2017 with 2020/2025/2030 benchmarks: from search results over Asia Society, a Interventions journal article ("This is China's Sputnik Moment": The Politics and Poetics of Artificial Intelligence, 2021), and summaries of AI Superpowers. The book was not read.
• Marc Andreessen's "AI's Sputnik moment", posted 27 January 2025 after R1's release on 20 January 2025: from contemporaneous reporting (NPR, 28 January 2025; Fortune; FDD, 30 January 2025). The original post was not retrieved.
• Alex Tabarrok, "The Sputnik vs. DeepSeek Moment: Why the Difference?", Marginal Revolution, 11 July 2025, read directly. Source for the quoted sentence and for the proposed NIH −40% / NSF −56% figures as he states them.
• Ellie Hardegree, "Reverse Sputnik: Is the US Squandering its Future in Science and Technology?", Stimson Center, 25 April 2026. The page returned HTTP 403 and was not read; the title, author, date and thesis are from search-result summaries only, and nothing is quoted from it.
• NSF FY2026 outcomes — the −56.9% request, the roughly 5% enacted cut to non-defence R&D, the $8.75 billion appropriation signed 23 January 2026, and the 613 grants at about 20% of the FY2021–24 pace reported by APS News in April 2026 — are partly carried from arpanet-1969's CRS-sourced section and partly from search results. The $42 billion / 21% proposed cut is from Science and CRA coverage of the FY26 budget request. The enacted appropriations text was not read.
• The Genesis Mission executive order of 24 November 2025 and the $150 million through September 2026 under OBBBA § 50404: from law-firm client alerts (Morgan Lewis, Holland & Knight, Morrison Foerster, December 2025), DOE announcements, and the Institute for Law & AI. The executive order text at whitehouse.gov was not read directly.
• Hyperscaler 2026 capital expenditure of $700–800 billion is carried from arpanet-1969, where it is flagged as an analyst aggregation rather than a filing. It is used here only as an order of magnitude and should not be quoted as a figure.
• Brian Weeden, "Bad Idea: Declaring a 'Sputnik Moment' (Unless You Really Mean It)", CSIS Defense360, 17 December 2021, read directly. Source for both quoted passages and for the corrected definition.
• Mark Hlavacik, "Do 'sputnik moments' spur educational reform? A rhetoric scholar weighs in", The Conversation, 3 October 2023, read directly. Source for the ARPA-for-education failure, the "easily forgotten" quote, and the A Nation at Risk (1983) and 2016 comparisons.
• The catalogue of declared Sputnik moments — Romney (May 2005), Napolitano and Pawlenty (April 2007), Obama's State of the Union of 25 January 2011 — is from U.S. News (26 January 2011) and Space.com via search summary; the Obama line is verifiable in the White House transcript archived at obamawhitehouse.archives.gov, which was located but not fetched.

What I could not establish. Whether the ARPA founding appropriation of $520 million is an authorisation, an obligation or a first-year outlay — the figure circulates without a source. Whether Eisenhower's "not one iota" sentence exists in any transcript. And the counterfactual that the whole entry turns on: what American computing funding would have looked like without the launch. The Soviet Union supplied one satellite; every claim about what it caused past the appropriations line is inference, including the ones made here.