Alpha School charges as much as $75,000 annually for a model built around two hours of AI-led academic work, promising students will learn twice as much as peers. Its TimeBack app personalizes instruction, while adult “guides” focus on motivation rather than lesson planning or grading; the rest of the day is devoted to activities such as public speaking and relationship-building. The model has attracted prominent technology and business supporters and expanded to about 1,500 students across roughly 50 campuses. Yet the school network’s own figures show only 42% met its goal of doubling their growth in both math and reading. Math is the sharper weakness: about half of K-8 students missed the target, compared with roughly one in five for reading.
Students who entered behind national benchmarks did especially poorly: among those at or below the 50th percentile in math, just 37% achieved the network’s growth target. Reports from parents and former employees raise further concerns, including anxiety, late-night work, and faulty AI-generated lessons. Alpha’s software is also being used by more than 2,600 public-school students in Texas and South Carolina, where districts face a different test because they cannot screen out struggling pupils. A Houston pilot reportedly took more student time than planned. Broader research adds caution: AI use can improve homework performance while weakening exam results, and frequent use has been associated with lower science scores. The central question is whether AI accelerates durable learning or merely helps students complete work.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.