“Super Intelligence” or the Enemy? Trump’s Latest Word Game Doesn’t Survive Contact with Reality
President Trump's recent declaration that anyone still using the term "artificial intelligence" rather than his preferred "super intelligence" is "THE ENEMY" of the White House is not merely eccentric branding. It is a small but revealing piece of nonsense that exceeds even the uncritical boosterism of Operation Warp Speed. Where Warp Speed at least involved an actual emergency technology with measurable clinical endpoints, the rebranding of current AI systems as "super intelligence" invents a capability that does not yet exist, collapses a useful technical distinction, and treats scepticism as disloyalty.
In technical and philosophical discussion, "artificial superintelligence" (ASI) has a relatively clear meaning: an intellect that greatly exceeds the cognitive performance of humans across virtually all domains of interest. It is a hypothetical future state, not a description of today's large language models, diffusion systems, or agentic tools. Those systems are narrow, brittle, data-hungry, prone to confident fabrication, and dependent on enormous amounts of human-generated training material and ongoing human oversight. Calling them "super intelligence" does not make them superior to human judgment in the broad sense the term implies; it merely papers over the gap between demonstrated capability and marketing aspiration. The executive order directing federal agencies to stop acknowledging "artificial intelligence" or "AI" in official communications does not alter that technical reality. It simply requires government documents to use a more flattering synonym for the same systems already in use.
The "enemy" formulation makes the exercise worse. Language is not a loyalty test. Reasonable people can prefer the established term "AI" because it accurately signals that the systems are artifacts of human design rather than autonomous minds. Others can note that current models fall well short of the performance thresholds associated with superintelligence in the literature. Treating either position as enmity is the sort of rhetorical escalation that substitutes status and personal branding for analysis. It also sits awkwardly with the continued use of "AI" on some White House and agency pages even after the order, an inconsistency that suggests the rebrand is more performative than operational.
Compare this with Operation Warp Speed. That effort involved compressed timelines, massive public funding, and regulatory flexibility for vaccines against a novel pathogen. It produced products even while legitimate questions about duration of protection, transmission effects, adverse-event surveillance, and the wisdom of broad mandates remained open. The process was imperfect and, in places, overconfident, but it was tethered to empirical benchmarks. The "super intelligence" push has no comparable tether. There is no public trial protocol that demonstrates current systems exceed human intelligence across domains, no independent verification regime that would falsify the claim, and no serious engagement with the well-documented failure modes, hallucination, goal misgeneralisation, concentration of capability in a handful of firms, energy and data demands, and the risk of deploying systems whose behaviour cannot be fully predicted. Instead, the administration has hosted industry leaders, accepted a voluntary "morally binding" accord with limited enforcement teeth, and directed agencies to adopt more positive language.
An AI-critical perspective does not require denying that the underlying techniques can be useful for specific tasks. It does require resisting the substitution of nomenclature for capability, and the conversion of caution into political opposition. The practical risks: labour displacement without corresponding productivity gains that reach most workers, automated decision systems that scale errors, surveillance applications, and the concentration of power among the firms that control the models and the infrastructure, are already visible. Labelling the technology "super" in advance of those risks being managed does not reduce them; it encourages the opposite posture of treating further deployment as inevitable and scepticism as hostile. Words can shape expectations, but they cannot create the missing competence or the missing accountability. Current systems remain artificial in the precise sense that their intelligence is engineered, limited, and contingent. Pretending otherwise, and calling the pretence mandatory, is a weaker form of technological faith than Warp Speed ever was.
So, just who was whispering in Trump's ear this time?
