AI Is Not Insight
AI can accelerate access to information and automate processes, but insight still depends on judgment, context, curiosity, and the ability to ask meaningful questions. For higher education, the challenge is not simply technological adoption, but institutional adaptation.
Across higher education and education-adjacent organizations, AI has moved from experimentation to expectation. Job postings now routinely foreground AI skills. These skills range from prompt engineering to “vibecoding.” Many roles now demand these skills, even though they had no explicit technical need until recently.
This shift signals something real. Nonetheless, it is also creating a familiar risk. We adopt tools faster than we clarify the problems they are meant to solve.
The signal in this moment is not the proliferation of tools. It is the expectation that individuals and organizations can move more quickly from question to insight. The noise is the assumption that tools alone will get us there.
Most education organizations do not lack data. They lack usable insight.
They produce dashboards that are rarely consulted. Their reports arrive too late to influence decisions. They create evaluation frameworks that run parallel to, rather than within, how the organization actually functions. The result is a persistent gap between information and action.
AI has the potential to close that gap. But only if we are clear about what is actually broken.
From Reporting to Decision Systems
Traditional evaluation models are built for reporting:
- What happened
- How a program performed
- Whether targets were met
These outputs are necessary. They are also insufficient.
They are retrospective, episodic, and often disconnected from the decisions they are meant to inform.
What organizations need instead are decision systems:
- What is happening now
- What is likely to happen next
- What should we do about it
That shift, from reporting to decision support, is not new. What is new is the ability to operationalize it.
AI reduces the friction between question and analysis. It allows organizations to synthesize data across systems, generate analyses on demand, and test assumptions quickly. But speed alone does not create value. Value comes from embedding analysis into the decision process itself.
The Real Constraint: Structure, Not Data
In building institutional research and evaluation systems, the limiting factor is almost never access to data. It is structure.
- Data definitions vary across units
- Ownership is fragmented
- Systems are poorly integrated
- Evaluation frameworks are not aligned with operational priorities
- Outputs are not designed for how decisions are actually made
AI does not solve these problems. It accelerates them.
If the underlying data is inconsistent, AI produces faster inconsistency.
If the evaluation model is unclear, AI produces more noise at scale.
The rapid rise of AI-related hiring often emphasizes tools over underlying data practices. This suggests we are scaling capability faster than we are strengthening the systems on which those capabilities depend.
The prerequisite for effective AI use is not more data. It is disciplined structure: clear definitions, aligned metrics, and systems that reflect how the organization actually operates.
Without that, AI becomes a multiplier of confusion.
AI as an Accelerator of Thinking
Where AI does create value is in how analysis happens.
Traditional workflows are linear:
- extract
- clean
- analyze
- report
They are also slow and rigid, which limits how many questions an organization can realistically explore.
AI introduces iteration.
- Ask a question
- Generate an initial analysis
- Refine the question
- Test assumptions
- Reframe the output
This creates a tighter loop between thinking and analysis.
Technical fluency, including the ability to rapidly prototype analyses through AI-assisted coding (“vibecoding”), is becoming table stakes. But without clear data definitions and evaluation frameworks, faster prototyping simply accelerates unclear thinking.
In practice, this changes the nature of the work. Analysis becomes less about executing predefined steps and more about navigating a sequence of increasingly better questions. In my own work, this has meant:
- generating and refining analytical code and queries
- structuring evaluation frameworks and metrics in real time
- synthesizing qualitative and quantitative inputs into coherent models
- translating complex findings into decision-relevant outputs quickly
The advantage is not just efficiency. It is cognitive: the ability to explore more possibilities, challenge assumptions, and arrive at better answers.
What Changes for Education Organizations
If we take this shift seriously, several implications follow.
1. Evaluation becomes continuous, not episodic: Periodic reporting cycles give way to ongoing analysis embedded in operational workflows.
2. Data work moves closer to decision-making: The separation between those generating insights and those acting on them becomes a liability. Effective organizations collapse that distance.
3. Clarity becomes a core capability: As the volume of available data increases, the ability to produce clear and decision-relevant insight gains value. This insight becomes more valuable than the data itself.
4. Speed exposes weak assumptions: Faster analysis does not just accelerate answers. It exposes where the underlying questions, or models, are flawed.
The Work Ahead
AI will not replace evaluation. It will make its weaknesses visible.
Organizations that treat AI as a layer on top of existing reporting structures will see incremental gains. They will have faster reports. Their outputs will be more polished. Efficiency will improve slightly.
Organizations that use AI to rethink how data informs decisions will see something different. They will experience faster learning cycles, tighter alignment between strategy and execution, and more effective use of resources.
This is not primarily a technology shift. It is an operating model shift.
The organizations that benefit most from this moment will not be those that adopt AI the fastest. They will be those that align tools, data, and decision-making into a coherent system.
The shift underway is not about AI capability. It is about organizational clarity. And that is a harder problem, one that technology alone will not solve.
READ MORE
Essays on Higher Education, Research, Innovation, and Public Impact
Ongoing writing and analysis on higher education, research, innovation, institutional adaptation, and public impact.
Let’s Connect
I welcome opportunities to connect around higher education, research, innovation, student success, and cross-sector collaboration. Whether related to institutional strategy, public engagement, research partnerships, or applied innovation, I’m always interested in thoughtful conversations and meaningful work.