Sensemaking is the continuous interpretive process by which people and organizations figure out what they are doing and why, under conditions of ambiguity where multiple plausible readings of the situation coexist. It is not decision-making, which presupposes that options and criteria are already defined. It is the upstream activity through which the situation becomes intelligible enough that decision-making becomes possible at all. Weick identified seven properties: sensemaking is grounded in identity, retrospective, enactive, social, ongoing, focused on extracted cues, and driven by plausibility rather than accuracy. Each property overturns a commonsense assumption about organizational cognition. Together they describe how interpretation actually proceeds in the face of equivocality — not through the orderly survey of alternatives that management textbooks describe, but through the messy, iterative collision of partial interpretations with unfolding events.
There is a parallel reading that begins not with the social construction of meaning but with the material conditions that enable sensemaking to occur at all. Organizations don't merely "figure out what they are doing" through collective interpretation — they depend on vast infrastructures of documentation, communication channels, meeting spaces, and technological systems that shape what can be sensed in the first place. The seven properties Weick identifies operate atop a substrate of capital allocation, technical standards, and power relations that determine whose interpretations matter and which cues become visible. When a hospital misreads a situation catastrophically, the failure isn't just premature closure but the prior decisions about staffing ratios, documentation systems, and hierarchical structures that filtered what could be perceived.
The AI transition makes this infrastructure visible by disrupting it. The "imagination-to-artifact ratio" collapses not because sensemaking speeds up but because the means of sensemaking concentrate in platforms controlled by a handful of corporations. The prototype that appears before debate isn't just faster enactment — it's enactment through systems whose interpretive frameworks are embedded in training data, model architectures, and optimization functions invisible to users. The "twenty-fold productivity multiplier" measures not enhanced sensemaking but its outsourcing to black-boxed systems. Organizations still need to make sense of ambiguous situations, but increasingly they do so through tools that pre-structure what counts as a legitimate interpretation. The real transformation isn't from slow to fast sensemaking but from distributed human interpretation to centralized machine mediation, where the cues extracted and the plausibility assessed depend on computational resources most organizations neither own nor understand.
The canonical statement of the framework is Sensemaking in Organizations (1995), though the concept had been developing across Weick's earlier work on organizing as a process rather than a structure. The move was radical: where organizational theory had treated organizations as entities that exist and then act, Weick treated them as ongoing accomplishments that exist only insofar as sensemaking continues. Stop the sensemaking and the organization dissolves into a collection of individuals with incompatible interpretations.
The framework reframes every classical organizational question. Decision-making becomes downstream of interpretation. Strategy becomes the retrospective articulation of patterns already enacted. Leadership becomes the shaping of the cues that collective attention extracts from a saturated environment. The Tenerife disaster, the Mann Gulch fire, and the Bristol Royal Infirmary tragedy each became paradigmatic cases through which Weick demonstrated that the failure mode is almost never insufficient information. It is premature closure — the commitment to an interpretation so coherent that contradictory cues cannot break through.
For the AI transition, the framework is diagnostic in a way that efficiency metrics cannot be. The imagination-to-artifact ratio's collapse is not merely a production gain; it is a compression of the interpretive process that once separated idea from artifact. What the twenty-fold productivity multiplier measures is speed of enactment. What it does not measure is the sensemaking that the speed displaces — the debate that never happens because the prototype already exists.
Weick's framework is itself an act of sensemaking, and he held it loosely. The seven properties are not a checklist but a diagnostic lens. Adequate sensemaking is not true sensemaking; it is sensemaking rich enough to enable wise action while remaining alert to the cues the current interpretation does not explain.
Weick developed the sensemaking framework across three decades at the University of Michigan's Ross School, building on his 1969 reframing of organizations as processes of organizing rather than static structures. The 1995 book consolidated the seven-property model; subsequent work with Kathleen Sutcliffe extended it into organizational mindfulness and high-reliability organizations.
Interpretation precedes decision. The rational model's survey-of-options presupposes that the situation is already intelligible; sensemaking is the upstream labor that produces the intelligibility.
Seven properties, not a sequence. Identity, retrospect, enactment, sociality, ongoingness, extracted cues, and plausibility operate simultaneously, not in stages.
Equivocality differs from uncertainty. Uncertainty is missing data; equivocality is missing frameworks. More information resolves the first. Only more interpretation resolves the second.
Plausibility enables action; accuracy resolves later. The map need not be correct; it need only be sufficient to get the organization moving, because movement generates the information the map cannot.
Sensemaking can collapse. Under extreme conditions — novelty, speed, stress — the interpretive process fails before action does, and the resulting behavior looks like panic, rigidity, or paralysis depending on what identity remains.
Critics have argued that the framework's emphasis on plausibility risks endorsing post-hoc rationalization and that its retrospective character makes it difficult to distinguish genuine sensemaking from self-serving narrative construction. Weick's response — that the quality of sensemaking depends on the quality of cue extraction and the willingness to revise in light of new evidence — is itself contested, because the mechanisms for enforcing that quality are social and therefore vulnerable to the very dynamics the framework describes.
The question of what enables sensemaking yields different answers at different scales. At the immediate interpersonal level, Weick's framework dominates (90/10) — people do construct meaning through ongoing social interpretation, extracting cues and building plausible narratives. The seven properties accurately describe how a team meeting unfolds or how colleagues make sense of a reorganization. But zoom out to the organizational level and the material conditions become equally important (50/50) — the interpretive process depends on communication infrastructure, documentation systems, and the political economy of whose voice carries weight. The Tenerife disaster wasn't just premature closure but also the technological constraints of 1970s radio systems and the hierarchical dynamics of aviation authority.
At the societal scale where AI operates, the infrastructure view gains ground (30/70). The concentration of computational power and training data in a few corporations means that sensemaking increasingly happens through their mediation. Yet even here, Weick's insight remains partially valid — organizations still engage in collective interpretation, but now with AI as a participant whose contributions shape the extracted cues and plausible narratives. The "imagination-to-artifact ratio" collapse represents both phenomena simultaneously: genuinely faster enactment AND a shift in interpretive authority from human discussion to computational generation.
The synthetic frame that holds both views recognizes sensemaking as a layered phenomenon. The social-interpretive layer that Weick mapped remains real and vital, but it operates atop infrastructural layers that shape what can be sensed and made sense of. AI doesn't replace sensemaking but restructures its substrate — changing not whether organizations interpret ambiguity but through what means and with what constraints. The framework needs both lenses: Weick's to understand how interpretation proceeds, and the infrastructural view to understand what enables and limits that interpretation.