No Country Can Buy What America's 3,900 Colleges Already Have

This article was originally published in Forbes on August 31, 2026.

The United States has roughly 3,900 degree-granting colleges and universities. Most of them are running a live experiment in what artificial intelligence does to teaching, to assessment and to the worth of a credential. Almost none of them can see what the others have found.

What would this country have if they could?

The experiment is already underway

The raw material is in place and already paid for. No ministry designed this arrangement. Thousands of institutions, several thousand disciplines, students of every age, all testing the same technology under real conditions in one academic year. No national system could manufacture that variety on purpose, and no amount of money buys it from a vendor. The U.S. possesses it, but what it has never built is a way for one institution's discovery to reach the next before the semester ends.

The urgency is quantifiable. Microsoft's 2026 AI in Education Report, a survey of 3,345 students, educators and leaders across six countries, found:

88% of educators and 92% of students and education leaders are already using AI for school-related work, while 77% of students reported no formal training in it.

Adoption did not wait for policy, and it will not wait for consensus.

What systemness makes possible

I serve in an advisory capacity on an AI working group in the California State University system, and watching that system operate changed how I think about scale. CSU is the largest four-year public university system in the country, with 22 universities, more than 470,000 students, and over 61,000 faculty and staff. Systemness, the word people inside use, means working something out once and carrying it to every campus. I see firsthand what happens when insights get gathered and the best of them get passed around. A student finds help at 11 at night and keeps going. A professor with 20 years of expertise discovers new superpowers. An adult returning to school reaches the degree sooner because the path was made clearer. Each of those began on one campus and reached 22. What could this country do if that multiplier ran across thousands of institutions?

The more impressive accomplishment is that each university kept its own judgment while the system moved together. A campus in the Central Valley and one in the aerospace corridor need different things from identical tools, and each retained the authority to decide what. Common capability underneath, local decisions on top. That combination is considerably harder to construct than either extreme, and it is the pattern worth exporting.

Build the road between the edges

I spent 20 years inside a Fortune 100, establishing Cisco's global innovation centers across 14 countries, most of them on university campuses. The most durable lesson from that work is that good ideas originate at the edges and stay stranded there unless somebody builds the road between them. Centralize too aggressively and the edges stop trying. Leave them alone entirely and each one pays full price to solve an identical problem. Value accumulates in the connective work between those two failures, and that work is the cheapest item on the list.

America has 3,900 laboratories. What it needs is the wiring. Three things determine whether that wiring gets built.

1. An exchange for what already works. The fastest way to lift a system is to move what one campus proved to the one that has not tried it. EDUCAUSE reaches more than 2,100 organizations, the American Council on Education represents nearly 1,600 institutions, and the National Science Foundation funds a coordination network for it. The channels are built, and every institution that adds to them makes them stronger. What did your institution learn this year that another one could use tomorrow, and who has heard about it?

2. Faculty as the country's applied research corps. The instructors rebuilding their own courses around these tools are producing the most useful evidence anyone has about AI and learning, usually on their own time and without recognition. Most of that evidence never leaves the classroom where it was made. Whose promotion file counts that work, and whose does not?

3. Skeptics as the sharpest reviewers. Concerns about assessment, academic integrity and hard-won expertise are frequently well founded. Institutions that hear them early write the policies everyone else copies two years later. Where does that conversation happen on your campus, and does anything change after it?

The lesson of the Morrill and Smith-Lever Acts

America has run this pattern before. The Morrill Act of 1862 established no national university. It transferred federal land to the states and let them build 57 institutions to their own specifications, teaching agriculture and engineering to people who had been shut out of higher education.

Those campuses spent five decades generating research that mostly stayed on campus. Then the Smith-Lever Act of 1914 supplied the connective layer, an extension service carrying findings from the laboratory bench to the county fairground. The experiments came first, the wiring came second, and it turned a scattered collection of colleges into a capability that fed a century of American growth.

We are living between those two acts, and this is the most consequential experiment American education has ever run, one we cannot afford to fail. The instruction that makes it work is old and simple. Think globally and act locally.

If enough of these experiments get shared instead of stored, the return is a longer working life for people whose skills would otherwise expire on schedule: fewer forced exits at 52, cheaper transitions, steadier towns. Call it the relevance dividend. It accumulates institution by institution, in curriculum committees and meetings that never make headlines.

That holds wherever you sit. The capability is national and every decision that builds it is local. Each institution that shares what it learned shortens the road behind it, and a country that gets this right hands a generation the ability to stay relevant on purpose, and gives the world a model it can borrow instead of buy.

Here is what makes Alex a credible voice on this topic: Alex serves in an advisory capacity on an AI working group in the California State University system, the largest four-year public university system in the country. Before that, he spent 20 years establishing Cisco's global innovation centers across 14 countries, most of them on university campuses.

Working out what AI systemness looks like for your campus or system? book a conversation →