Pith. sign in

REVIEW 3 major objections 4 minor 2 references

Deep Hype in Artificial General Intelligence: Uncertainty, Sociotechnical Fictions and the Governance of AI Futures

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read AGI's lasting hype is not an accident but a recursive loop: venture capital speculation, uncertainty, and sociotechnical fictions feed one another, keeping AGI permanently out of reach while steering AI governance.

desk verdict Deep hype is a genuinely useful concept for long-cycle technological promises, but the paper's 'not technically feasible' premise is asserted against its own cited sources and the corpus is too vague to fully check. read the letter →

arxiv 2508.19749 v2 pith:77UAZYTT submitted 2025-08-27 cs.CY

classification cs.CY
keywords ArtificialGeneralIntelligenceDeepHypeSociotechnicalFictionsUncertaintyVentureCapitalCyberlibertarianismLongtermismGovernanceofAIFutures
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to explain why Artificial General Intelligence keeps attracting record-breaking investment and political attention despite having no agreed definition and no demonstrated feasibility. The answer it proposes is 'deep hype': a long-term, overpromissory dynamic in which a network of uncertainties — what AGI is, when it will arrive, what it will do, what it is worth, what it risks — makes the promise nearly impossible to verify in the present, and thus keeps the window of investment open. The uncertainties are articulated through 'sociotechnical fictions': speculative visions presented with technoscientific legitimacy as if they were facts. The paper argues that venture capital logic converges technological and financial speculation, establishing a recursive relation: fictions tame and excite uncertainty, uncertainty feeds hype, and hype feeds the fictions. A reader should care because this reframes AGI's inflated promise as an engineered dynamic that narrows political debate and advances a cyberlibertarian, longtermist programme in which private actors, not democratic institutions, steer the technological future.

What carries the argument

Two concepts carry the argument. 'Deep hype' is a long-term, overpromissory dynamic that constructs visions of civilisational transformation through a network of uncertainties extending into an undefined future, making promises nearly impossible to verify in the present — unlike conventional hype's short cycles. 'Sociotechnical fictions' are mediated forms of imagination produced within science and technology, sheltered by technoscientific legitimacy and so not recognised as fiction; they expand on the idea of fictional expectations and perform epistemic, aesthetic, affective, and normative work — rendering imaginary futures tangible, generating wonder and fear, and scripting collective acti

What would settle it

A verifiable demonstration of general intelligence — a system passing independently audited, pre-registered benchmarks for cross-domain learning and adaptation — would falsify the claim that AGI is a sociotechnical fiction rather than an engineering target. Short of that, a replicable corpus study with explicit selection rules showing that AGI discourse does not actually revolve around the nine uncertainty arenas, or evidence that funding follows demonstrated capability rather than promissory narrative, would undercut the recursive-loop mechanism.

Watch

Extended reading notes

Core claim

On its own terms, the central claim is that AGI deep hype is the result of venture capital logic that converges technological and financial speculation, and that the sociotechnical fiction of AGI establishes a recursive relation to deep hype: hype is sustained by uncertainties that sociotechnical fictions tame and excite, while the fictions are fed by the hype and uncertainty they produce. AGI is presented as the most prominent contemporary sociotechnical fiction — an imagined entity sheltered by technoscientific legitimacy and treated as an inevitable fact. The paper maps nine arenas of uncertainty (conceptual, application, temporal, value, governance, economic, geopolitical, existential, e

Load-bearing premise

The diagnosis depends on AGI being technically infeasible — the paper states that AGI functions as a sociotechnical fiction precisely because it is not technically feasible — and if AGI turns out to be on a plausible path to being built, the line between fiction and ordinary promissory research blurs and the central claim weakens.

Editorial extensions

If this is right

  • AGI's vague definition and shifting timelines are structural features, not failures: they keep the promissory framework open and make the hype unfalsifiable in the present, which is exactly why investors and promoters cultivate ambiguity.
  • Regulatory and public attention that responds to AGI's promises rather than to demonstrated capabilities plays into the loop, because existential-risk narratives are themselves sociotechnical fictions that legitimise the actors who raise the alarms.
  • AGI hype functions as a form of soft anticipatory governance: it narrows political debate, frames the AGI future as unavoidable, steers research and investment, and consolidates the cyberlibertarian and longtermist agenda while displacing near-term concerns such as bias, environmental impact, and labour displacement.
  • Because hype and sociotechnical fictions remain reactivatable, AGI hype should not be expected to follow the classic hype-cycle collapse; instead, expectations can be renewed and restaged after any disappointment.
  • The same deep-hype dynamic should be visible in other perpetually imminent technologies — the paper names nuclear fusion, quantum computing, and brain-computer interfaces — wherever civilisational promises meet unresolved uncertainties and speculative capital.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Testable prediction the paper leaves implicit: each time a concrete AI milestone is reached, the goalposts will move rather than AGI being declared achieved, because a defined and verified AGI would collapse the fiction that sustains current valuations. A timeline study of how milestones are relabelled (e.g., as 'narrow AI' or 'tools') would test this.
  • The nine uncertainty arenas could serve as an analytic checklist for other speculative technologies; if fusion or quantum discourse shows the same recursive pattern, that would generalise deep hype beyond AI; if not, the pattern may be specific to AGI's particular combination of vagueness and financial scale.
  • Because the paper draws on an unspecified series of statements and documents, the cleanest way to pressure its diagnosis is a replicable discourse analysis: if a systematically sampled corpus of AGI communications is not dominated by these uncertainty arenas, the recursive-loop claim would need revision.
  • A governance countermeasure follows from the paper's logic though the author does not state it: anchoring regulation to demonstrated, verifiable capability rather than to promised futures would break the loop at its weakest point, since the loop depends on promises being unverifiable in the present.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. This paper argues that Artificial General Intelligence (AGI) is sustained by a distinct form of hype—'deep hype'—defined in §4 as a long-term, overpromissory dynamic that constructs civilisational-transformation promises through a network of uncertainties extending into an undefined future, making verification nearly impossible in the present. Drawing on an unspecified 'series of statements and documents' from AGI leaders, the paper proposes nine uncertainty arenas (conceptual, application, temporal, value; governance, economic, geopolitical, existential, ethical) and claims that these uncertainties are articulated through 'sociotechnical fictions' that render not-yet-existing technologies intelligible, desirable, and urgent. The central conclusion is that AGI deep hype results from venture-capital logic converging technological and financial speculation, forming a recursive relation between uncertainty, fiction, and hype; this dynamic advances a cyberlibertarian and longtermist programme, sidelines democratic oversight, and functions as a form of governance of AI futures. The paper is framed as a theoretical contribution to hype studies and critical governance studies.

Significance. If the argument holds, the paper makes a useful contribution by moving hype analysis from short product cycles to long-term promissory dynamics and by explicitly linking hype to venture-capital speculation, sociotechnical imaginaries, and governance. It usefully synthesises hype studies, STS, and critical political economy, and it assembles a broad set of contradictory public statements by AGI leaders. The concept of 'sociotechnical fiction' is a potentially valuable extension of Beckert's fictional expectations. The paper's main weaknesses are that the empirical corpus is not specified, and that the load-bearing premise that AGI is 'not technically feasible' is asserted rather than established, with the paper's own cited sources (Kaplan et al. 2020; Bubeck et al. 2023; Agüera y Arcas & Norvig 2023) cutting the other way. The recursive 'deep hype' mechanism is theoretically coherent but risks circularity. With targeted revisions to the infeasibility claim and a transparent account of the corpus, the paper could be a valuable critical intervention.

major comments (3)
  1. [Discussion – The role of Sociotechnical Fictions in AGI deep hype] The argument that AGI is a sociotechnical fiction—and the paper's headline claim—rests on the sentence: 'AGI functions as a sociotechnical fiction precisely because it is not technically feasible.' This premise is load-bearing and is not defended. The paper itself cites Kaplan et al. (2020) scaling laws as a 'proof of verifiable truth' (Application uncertainty section), Bubeck et al. (2023) on 'sparks' of AGI (Temporal uncertainty section), and Agüera y Arcas & Norvig (2023) claiming LLMs already embody AGI. These sources undercut the asserted infeasibility. If AGI is a contested but potentially feasible technology, the distinction between sociotechnical fiction and ordinary promissory research collapses, and the conclusion that AGI is 'the most prominent contemporary sociotechnical fiction' becomes unsupported. Please either engage substantively with the cited feasibility claims (with t
  2. [The Uncertainties of Artificial General Intelligence (Section 4)] The paper repeatedly states that its analysis is 'based on the analysis of a series of statements and documents from key AGI actors,' but it never specifies the corpus: no selection criteria, source list, time period, number of documents, sampling strategy, or search procedure. This makes the nine-uncertainty-arena taxonomy and the fit between evidence and framework uncheckable. The reader cannot tell whether the quotes were selected to illustrate pre-existing categories or whether the categories emerged from a systematic corpus. If the paper is intended as a theoretical essay, the empirical material should be explicitly labelled as illustrative rather than evidential; if it is intended as an empirical discourse analysis, a methodological appendix is required. As written, the empirical grounding is not auditable, which is a significant limitation for a paper whose central claims are abou
  3. [Definition of deep hype (Section 4) and recursive claim (Discussion)] The concept 'deep hype' is defined with the very features that the paper then discovers in AGI: long-term civilisational promises, a network of uncertainties, unverifiability, and a recursive relationship between hype and fiction. The nine arenas are generated from the same selected corpus, and the recursive relation is asserted in the Discussion rather than demonstrated. This is a circularity risk: the empirical sections can only confirm what was already built into the definition. The paper should explicitly acknowledge that 'deep hype' is an interpretive concept constructed through this case, not a hypothesis tested against it, or it should provide an independent operationalisation (e.g., distinguishing deep hype from other long-term promissory technologies such as nuclear fusion or quantum computing before applying it to AGI). Without this, the central contribution risks being an elab
minor comments (4)
  1. [Governance uncertainty] Chronology error: the paper says 'Six days after the hearing, on March 22 of the same year, the Future of Life Institute published an open letter...' The Senate hearing was on May 16, 2023; the FLI letter was March 22, 2023, before the hearing. Please correct the chronology.
  2. [Discussion – AGI Deep Hype] Typographical errors: a stray 'Te.' in the paragraph beginning 'Like nuclear fusion...'; 'Andreesseen' for Andreessen; 'Wachawami' for Wadhwani; 'an the Open University' in footnote 1; and the placeholder citation 'AUTORb' in the section on sociotechnical fictions. These should be cleaned up.
  3. [Discussion – AGI Deep Hype] The statement 'AGI is not rooted in empirical scientific progress but in speculative thought, financial interests and technopolitical agendas' is stronger than the evidence presented, especially given the paper's own references to scaling laws and the claimed 'sparks' of AGI. Consider softening to 'not primarily rooted' or providing supporting argumentation.
  4. [General] The list of nine uncertainty arenas is one of the paper's main analytic outputs, but it is presented as a simple enumeration. A table summarising each arena, its key actors, and representative quotes would improve readability and make the analytical structure easier to follow.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is a conceptual application, not a derivation, and its self-citations are attributions rather than load-bearing external proofs.

full rationale

The paper introduces 'deep hype' and 'sociotechnical fictions' as theoretical concepts and then uses them to interpret AGI discourse. This is definitional application rather than a derivation: there are no equations, fitted parameters, or first-principles predictions that collapse into inputs. The self-citation to Belsunces (2025) for 'sociotechnical fictions' is the author's own prior work, but the paper substantively defines the concept in the text and does not use the citation as an external authority to forbid alternatives or to force the AGI conclusion; it is an attribution. The claim that AGI is a paradigmatic case of deep hype follows from matching AGI's observed features (long-term promises, civilisational stakes, unresolved uncertainties) to the paper's stated definition; this is the standard structure of a conceptual argument, not a circular inference. Similarly, the recursive relation between hype and fictions is a proposed feedback model, not a derivation of the conclusion from itself. The main epistemic weaknesses are the asserted infeasibility of AGI ('AGI functions as a sociotechnical fiction precisely because it is not technically feasible') and the unspecified corpus of statements; these are correctness/evidentiary concerns, not circularity. Therefore, no circular step can be exhibited, and the score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 2 invented entities

The paper's claims rest on a set of interpretive choices rather than measurements: which statements count as evidence, which taxonomy organizes them, and the prior assumption that AGI is infeasible. There are no numerical fits, but the conceptual 'parameters' listed above do the same work as fitted constants in a quantitative paper.

free parameters (3)
  • Selection of statements and documents
    The paper analyzes 'a series of statements and documents from key AGI actors' without specifying inclusion criteria, corpus endpoints, or exclusion rules; which quotes appear in the nine uncertainty arenas is chosen by hand, so the fit between evidence and the deep hype frame is not independently checkable.
  • Nine uncertainty arenas taxonomy
    The four constitutive and five consequential arenas are the author's own analytical grid; the paper fits each quote into pre-chosen categories, which helps guarantee the framework finds what it looks for.
  • Three senses of 'deepness'
    The claim that AGI hype is deep in exactly three senses (civilisational transformation, complex effects, deep time) is asserted, not derived; these dimensions are chosen after the case.
assumptions (4)
  • domain assumption AGI is not technically feasible, so it qualifies as fiction
    Discussion section states 'AGI functions as a sociotechnical fiction precisely because it is not technically feasible.' The paper treats contested infeasibility as settled despite citing researchers who claim LLMs already exhibit AGI.
  • domain assumption Uncertainty in technoscience is productive and strategically manageable
    Section 'Uncertainty and Ambiguity' leans on Nowotny to assert uncertainty is co-produced and weaponizable; this grounds the claim that AGI stakeholders instrumentally cultivate ambiguity.
  • domain assumption Hype is performative and politically strategic
    The paper follows van Lente et al. and Bareis in treating hype as an agent in power struggles; the analysis assumes statements by tech leaders are deliberate strategic moves, not sincere predictions.
  • ad hoc to paper Representativeness of selected statements
    The quotes from Altman, Andreessen, Goertzel et al. are treated as evidence of a coherent cyberlibertarian/longtermist programme without a stated selection or sampling protocol.
invented entities (2)
  • deep hype
    purpose: New concept for long-term, unverifiable, civilisational overpromising, offered as the paper's central analytical contribution
    Defined in this paper and applied to a single case chosen to fit; no falsifiable prediction or external benchmark distinguishes it from the prior concepts it draws on.
  • sociotechnical fictions (as a distinct category)
    purpose: Fictions produced inside science and technology but carrying technoscientific legitimacy; engine of the deep hype dynamic
    The concept is introduced in the author's own prior work (Belsunces 2025, referenced in text as AUTORb), and its application here is internal to the framework.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Deep Hype in Artificial General Intelligence: Uncertainty, Sociotechnical Fictions and the Governance of AI Futures." pith.science (2026). https://pith.science/paper/77UAZYTT

@misc{pith2026250819749,
  author       = {Pith},
  title        = {Pith review of: Deep Hype in Artificial General Intelligence: Uncertainty, Sociotechnical Fictions and the Governance of AI Futures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/77UAZYTT}},
  note         = {Machine review of arXiv:2508.19749}
}
read the original abstract

Artificial General Intelligence (AGI) is promoted by technology leaders and investors as a system capable of performing all human intellectual tasks, and potentially surpassing them. Despite its vague definition and uncertain feasibility, AGI has attracted major investment and political attention, fuelled by promises of civilisational transformation. This paper conceptualises AGI as sustained by deep hype: a long-term, overpromissory dynamic articulated through sociotechnical fictions that render not-yet-existing technologies desirable and urgent. The analysis highlights how uncertainty, fiction, and venture capital speculation interact to advance a cyberlibertarian and longtermist programme that sidelines democratic oversight and reframes regulation as obsolete, with critical implications for the governance of technological futures.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

2 extracted references · 2 linked inside Pith

  1. [1]

    Ahmed, S., Jaźwińska, K., Ahlawat, A., Winecoff, A., & Wang, M. (2024). Field-building and the epistemic culture of AI safety. First Monday . https://firstmonday.org/ojs/index.php/fm/article/view/13626 Altman, S. (2025a, June 10). The gentle singularity . Sam Altman’s Blog. https://blog.samaltman.com/the-gentle-singularity Altman, S. (2025b, February 9). ...

  2. [32]

    https://doi.org/10.51628/001c.94404 Jasanoff, S., & Kim, S.-H. (2015). Dreamscapes of modernity: Sociotechnical imaginaries and the fabrication of power . University of Chicago Press. Kaplan, J., McCandlish, S., Henighan, T., Brown, T. B., Chess, B., Child, R., Gray, S., Radford, A., Wu, J., & Amodei, D. (2020). Scaling laws for neural language models . a...

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.