Aristotle

Getting into college is not the only challenge. Students also need to arrive ready enough to succeed.

ACT reported that 30% of nearly 1.4 million ACT-tested 2025 graduates met three or more of the four ACT College Readiness Benchmarks.

That does not tell the whole story of a student. But it does point to a real preparation problem. Reading, math, science reasoning, and study habits build over time.

College and career readiness starts before senior year.

Source: ACT, Oct. 15, 2025.

AI literacy is no longer just for computer science students.

Stanford HAI's 2026 AI Index shows a complicated picture: many students are already using AI for schoolwork, while school policies and formal pathways are still uneven. One striking figure: only 6% of teachers said their school AI policies were clear.

Students need more than access to tools. They need judgment: when AI helps, when it is wrong, what to verify, and what they still need to learn themselves.

Source: Stanford HAI, 2026 AI Index Report.

America's public schools are projected to get smaller. That does not make the work simpler.

NCES projects U.S. public K-12 enrollment will decline from 49.6 million in fall 2022 to 46.9 million in fall 2031, about a 5% drop.

At the same time, schools still have to deal with learning recovery, staffing pressure, absenteeism, special support needs, and new AI expectations.

Fewer students does not mean fewer challenges. It means schools need more precise ways to support the students they serve.

Source: NCES Condition of Education.

The teacher staffing problem is not the same everywhere.

NCES reported that public schools filled 79% of teaching vacancies with fully certified teachers before the 2024-25 school year. But specialized areas like ESL/bilingual education, special education, computer science, and career/technical education were harder to fill.

High-poverty schools and schools with more than 75% students of color filled only 72% of vacancies with fully certified teachers.

The students who need the most specialized support are often in the places with the least staffing capacity.

Source: NCES School Pulse Panel / IES, Oct. 17, 2024.

Tutoring can be powerful. But many schools are still struggling to staff it.

NCES reported that schools filled 73% of non-teaching vacancies overall before the 2024-25 school year. Tutor vacancies were filled at 65%, lower than most instructional support roles.

That is the scaling problem: students need more feedback and practice than schools can always provide through staffing alone.

Good education technology should extend support capacity while respecting the value of teachers and tutors.

Source: NCES School Pulse Panel / IES, Oct. 17, 2024.

Learning recovery is happening, but not evenly.

NWEA studied more than 5 million students across 9,326 U.S. public schools. By fall 2024, about one in three schools had recovered in math or reading. Only about one in seven had recovered in both.

That means the average does not tell the whole story. Some schools are back to pre-pandemic achievement levels. Many are not.

Students need support that responds to their actual gaps, not just broad assumptions about what everyone missed.

Source: NWEA, Feb. 24, 2026.

Most high school students are already using generative AI for schoolwork.

College Board reported that use rose from 79% in January 2025 to 84% in May 2025. But school rules are still all over the place: bans, no-policy use, teacher-by-teacher rules, department rules, and encouragement in all classes.

Students need clearer expectations. When is AI allowed? When is it not? What counts as real thinking? What should a student be able to explain after using it?

The policy conversation is not just about tools. It is about making school expectations understandable.

Source: College Board, Oct. 6, 2025.

AI is already part of many students' schoolwork at home.

Pew found that 64% of U.S. teens use AI chatbots, and 54% use them for schoolwork. But only 51% of parents say their teen uses chatbots, and about 30% are unsure.

That does not mean parents are doing anything wrong. It means the technology moved faster than the family routines around it.

The useful next step is simple: ask students what they are using, when it helps, when it gets things wrong, and what work they still need to do themselves.

Source: Pew Research Center, Feb. 24, 2026.

A line we kept hearing for the past decade: American higher education was in structural decline. Enrollment peaked at 18.1M in 2010, fell every year through 2022, and bottomed at 15.4M. Two-point-seven million students gone.

That story is now wrong. Fall 2024 was up 3% year-over-year. Fall 2025 came in at 16.2M — the strongest single-year gain in over a decade.

The interesting part: this rebound is happening against the demographic cliff, not before it. Smaller cohorts of 18-year-olds start arriving on campus this year. Something other than demographics is bringing students back.

The mix matters too. Community-college enrollment is still down 38% from peak; four-year publics are up 15%. Who returns, and to what kind of institution, will shape what college means for the middle and high schoolers we work with — kids who are eighteen months to four years from making this call.

Source: National Student Clearinghouse Research Center, Current Term Enrollment Estimates (Fall 2025 preliminary); NCES IPEDS, Digest of Education Statistics Table 303.70. https://nscresearchcenter.org/current-term-enrollment-estimates/

Georgetown's Center on Education and the Workforce dropped *The Major Payoff* in October and one number stuck with us: the gap between the top-paying college major and the bottom is $3.4M over a lifetime. Bigger than the gap between getting a degree and not getting one.

Top of the list: petroleum engineering, computer engineering, chemical engineering. All between $115K and $146K median annual for prime-age workers. Bottom: early childhood education, social work, human services. Forty-nine to fifty-eight thousand — about what someone earns without a four-year degree.

Two things we keep thinking about. The bottom of the list is full of jobs that involve teaching small children. The market is pricing those. Whether the price is right is a different conversation than the one a 17-year-old picking a major can have.

The other: variance inside each major is enormous. The 25th-percentile engineer earns less than the 75th-percentile English major. The major matters. The student matters more.

The teens we work with are eighteen months from this choice. The most useful thing a tutor — human or AI — can do is help them get good enough at hard subjects early enough that the door to engineering, or finance, or research, stays open longer than it otherwise would. You can always close a door later. You can't always reopen one.

Source: Anthony P. Carnevale et al., *The Major Payoff*, Georgetown University Center on Education and the Workforce (October 2025). https://cew.georgetown.edu/cew-reports/major-payoff/