Joining their Big Ten brethren, Purdue University recently announced that they will be adding an “AI working competency” graduation requirement that will go into effect for first-year students entering Fall 2026.
I have some questions. And also some concerns.
Back in October I shared my impressions and experiences having travelled to a number of different institutions that are directly confronting the challenge of how to evolve instructions and operations to deal with the existence of generative AI technology.
I identified a number of common approaches that seemed to be bearing fruit when it comes to engaging and energizing the entire university community at meeting these challenges.
This is, above all, a chance to refresh and renew the work of the institution and in the places operating in that spirit I’ve witnessed considerable hope for a good future.
- Administrations must lead, but they should lead from a position of institutional values that are centered in the discussion.
- The challenge must be viewed as a collaboration between administration, faculty, staff and, yes, students, where each constituent group has an opportunity to articulate their views under the umbrella of those root institutional values.
- As part of this process, there must be space for difference that preserves individual freedoms. Faculty should both have the resources necessary to experiment with AI use and the power to refuse its integration. Students must ultimately be respected as the chief agents behind their own educations.
- The discussion must go beyond merely adding another untethered competency. As I say in that previous column, we must “do more than doing school.” Layering AI on top of the status quo is a missed opportunity to reimagine the work of teaching and learning in ways that will make institutions far more resilient to whatever additional technological changes are coming.
Working from the reporting at Forbes, I can declare with some confidence that what Purdue is proposing is the opposite of these emergent best practices that I have seen elsewhere.
Here is how the initiative is characterized:
The requirement will be embedded into every undergraduate program at Purdue, but it won’t be done in a “one-size-fits-all” manner. Instead, the Board is delegating authority to the provost, who will work with the deans of all the academic colleges to develop discipline-specific criteria and proficiency standards for the new campus-wide requirement. Chiang [Purdue President Mung Chiang] said students will have to demonstrate a working competence through projects that are tailored to the goals of individual programs. The intent is to not require students to take more credit hours, but to integrate the new AI expectation into existing academic requirements.
This is, in a word, impossible. At least it’s impossible in any way that’s genuinely meaningful or useful to students.
Purdue has over 40,000 undergraduate students. They have more than 200 majors. They offer thousands of different courses. They have thousands of faculty. The expectation is that specific proficiency standards will be created for every single one of these programs, and after that, students will have to be held accountable to these proficiencies by next Fall by doing “projects.”
Does that sound possible? Because it’s not.
As a source of comparison, consider how long it takes to redo an institution’s general education curriculum which involves many fewer courses and fewer faculty.
Consider, also, as should be clear to anyone paying attention, we “don’t know how to teach AI.”
The fact that we don’t know how to teach AI is why the institutions engaging in the best practices have provided resources to the people best placed to figure out what kinds of proficiencies, experiences and projects may be useful, the faculty. They are treating the problem seriously as a challenge for the university to figure out how to serve its constituents.
Purdue is offering a press release, not a plan. It’s not even a policy. They are cooking up a recipe for chaos and demoralization, for half-assed B.S. meant to satisfy a bureaucratic box-checking exercise. This is serious stuff, and Purdue is treating it unseriously.
It’s worth asking why. One reason may be that Purdue is well-enmeshed with their corporate partners (primarily Google) and locking in experiences that use the products of these partners is pleasing to those partners. Purdue Provost Patrick Wolfe said, “it was ‘absolutely imperative that a requirement like this is well informed by continual input from industry partners and employers more broadly,’ and therefore he has ‘asked that each of our academic colleges establishes a standing industry advisory board focusing on employers’ AI competency needs and that these boards are used to help ensure a continual, annual refresh of our AI curriculum and requirements to ensure that we keep our discipline-specific criteria continually current.”
But guess who else doesn’t know what AI competencies employees need? Employers!
And guess who else doesn’t know what we’re supposed to be doing with this stuff? The tech developers themselves! Microsoft recently “scaled back” its AI sales targets because “nobody is using copilot.”
Purdue is sending clear signals to both students and employers that they are in the business of producing certified meat widgets for the AI-mediated future, even as we have no idea what that future may entail. There may be some students who find this proposition attractive, but it is not a leap in logic to imagine that the endpoint of this future is one where large, costly entities like Purdue University are not a necessary part of the equation.
Also, I think there is significant evidence that AI meat widget is not what students are looking for from their university experiences.
I will lay down a marker and predict there will be much sound and fury to meet the demands of the board, but it will signify nothing.
There is no need to waste everyone’s time chasing phantoms. We know how to work this problem, and what Purdue is doing ain’t it.
