TESCREAL Framework (Gebru and Torres)

TESCREAL (pronounced tes-cree-all) is an acronym coined by Émile Torres and developed in collaboration with Timnit Gebru to describe a constellation of overlapping techno-utopian belief systems that, they argue, shape Silicon Valley's relationship to AI. It bundles seven movements — Transhumanism, Extropianism, Singularitarianism, Cosmism, Rationalism, Effective Altruism, and Longtermism. The framework's fullest statement is Gebru and Torres's peer-reviewed paper "The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence," published in First Monday 29(4) in April 2024 (The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence (Gebru & Torres, 2024)); it is most accessibly summarized in Heaven's MIT Technology Review feature (July 2024). It has become a common analytical lens for the ideological underpinnings of the AGI race, and is itself contested as a polemic.

The acronym

Gebru and Torres's critique

Gebru and Torres advance three interlocking claims:

1. Common tenets unite the bundle. All seven worldviews share: (a) a belief that an all-powerful technology (AGI/superintelligence) is within reach and inevitable; (b) a belief that this technology could "level up" humanity; and (c) a do-or-die racing logic ("if we don't build it, someone else will").
2. Intellectual lineage with eugenics. Gebru and Torres trace the bundle's "improving humanity" framing to early 20th-century eugenics movements, citing the same logic of bettering the species and the same Western and white assumptions about who counts.
3. Resource-allocation critique. Resources flow toward fantastical futures (life extension, planetary colonization, AGI) at the cost of present-day harms (labor exploitation, racial bias, environmental damage) that disproportionately affect billions of non-Western, non-white people.

The First Monday paper characterizes the seven movements as "interconnected and overlapping ideologies" that emerged from the twentieth-century Anglo-American eugenics tradition, and argues that discriminatory attitudes associated with that tradition remain widespread within the AGI movement (Source: firstmonday.org). It also advances an engineering-methodology claim distinct from the ideological genealogy: because "AGI" is an undefined system rather than one built for specific applications, the authors argue it cannot be appropriately tested for safety under standard engineering principles, and they urge researchers to instead build defined, well-scoped systems for which safety protocols can be developed (Source: firstmonday.org).

Application to the AGI race

Gebru and Torres present the framework as an analytical lens for several features of frontier AI. They use it to explain why frontier labs (OpenAI, Anthropic, Google DeepMind) all have AGI or superintelligence in their stated mission, and why the same labs simultaneously make racing claims ("if we don't, China will") and existential-risk claims ("but it might kill us all"). Altman's "It's all the sci-fi stuff" answer to "what keeps you up at night?" is offered as an illustration. In this reading doomerism and boosterism coexist as a single ideology rather than opposing camps, with Hendrycks, Hinton, Amodei, MacAskill, and Tegmark's Future of Life Institute placed at various points on the spectrum. The framework also names a resource-allocation pattern in which billions of dollars go to AGI safety research while comparatively little goes to labor displacement, environmental remediation, or near-term harm mitigation.

Debates and positions

The framework is contested on several grounds. Critics argue the acronym lumps distinct intellectual traditions into a single caricature, noting that few self-identified transhumanists, effective altruists, or rationalists endorse all seven labels simultaneously. The eugenics linkage is contested by historians who argue the threads are weaker than Gebru and Torres present. Defenders of the constituent movements argue that effective altruism as practiced funds many present-day interventions, such as anti-malarial bednets and cash transfers, and that longtermism is a research project rather than a religion.

TESCREAL has entered the vocabulary of AI policy criticism and is associated with the broader critique cluster of Gebru, Emily Bender, and Alex Hanna. It is sometimes positioned as a counterweight to capability-extrapolation arguments. Discussions of frontier labs, racing dynamics, AGI Timelines, Longtermism, Effective Altruism-adjacent figures, and the three theories of victory frequently note whether the TESCREAL frame applies to the position under discussion.

See also

• Longtermism — the L of TESCREAL
• AGI Timelines — the bundle's prediction structure
• Three Theories of Victory (US / EU / China AI Governance) — the do-or-die racing logic the bundle inherits
• AI Divides (Literacy / Occupational / Ethico-Philosophical) — the resource-allocation critique
• Stochastic Parrots (Bender, Gebru, McMillan-Major, Mitchell, 2021) and the Octopus Test — sister critique from the same authorial cluster

The paper's two separable claims

The First Monday paper advances an engineering claim and a genealogical claim, and they stand or fall independently. The engineering claim is that "unlike systems with specific applications which can be evaluated following standard engineering principles, undefined systems like 'AGI' cannot be appropriately tested for safety" — safety testing presupposes a specification of intended function, and a system defined as able to perform "any task under any circumstance" supplies none. This yields the paper's recommendation to work "on defined tasks for which we can develop safety protocols," and does not depend on the eugenics history at all.

The genealogical claim is the contested one: that the seven ideologies are "direct descendants of first-wave eugenics," emerged "in roughly this order," and were in many cases "shaped or founded by the same individuals," such that "many of the very same discriminatory attitudes that animated eugenicists in the past… remain widespread within the movement to build AGI." Its evidence is authorship and funding — the first AGI book co-authored by an Extropian aiming at "transhuman AGI," and billionaire AGI funding from individuals affiliated with the named movements — rather than doctrinal continuity.

The paper also grounds its case in present rather than speculative harm, cataloging worker exploitation, data theft, environmental racism, misinformation, plagiarism, and the amplification of hegemonic views as already-documented consequences of the AGI race.

On standpoint, the authors disclose rather than claim neutrality: one is an engineer with nearly two decades in the technology industry, and the other "was a noted contributor to, and advocate of, the TESCREAL movement for nearly a decade before leaving the movement" — the basis of the paper's primary-source access, and the ground on which critics question its neutrality (The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence (Gebru & Torres, 2024)).

Relationships

• related: Longtermism — the L of TESCREAL
• related: AGI Timelines — the bundle's prediction structure
• related: Three Theories of Victory (US / EU / China AI Governance) — the "do-or-die racing" logic the bundle inherits
• related: Timnit Gebru — co-developer of the framework
• related: Alex Hanna, Emily M. Bender — the "humanist" critique cluster the framework anchors
• related: AI Divides (Literacy / Occupational / Ethico-Philosophical) — the resource-allocation critique
• related: Stochastic Parrots (Bender, Gebru, McMillan-Major, Mitchell, 2021) and the Octopus Test — sister critique from the same authorial cluster
• supports: What is AI? — MIT Technology Review (Heaven, 2024) — primary expository source in the wiki