Log in Sign up
Wiki Serving a 3-Trillion-Parameter Model Takes 59 Wafers, and Cerebras Has Published the Arithmetic Business Micron's Taiwan, China Unions Want 15 Percent of Operating Profit. One Plant Settled, One Walked Out AI Agents MCP Adds Durable Tasks and Mid-Job Questions, Handing Server Authors a Duplicate-Charge Problem News Only 42 Percent of Alpha School Students Hit the 2x Growth Its AI Model Promises, New York Times Finds Opinion AI Made Junior Patent Lawyers Instantly Better. Three Months Later, Only the Seniors Had Kept the Gains News TikTok's US Joint Venture Is Hunting Four Gigawatts of Computing Power by the Early 2030s Business Google Expands Its Virtual Power Plant Tenfold to 1,070 Megawatts, Paying Homes and Businesses to Step Aside for AI Data Centers Security Inside Meta's Law Enforcement Outreach Team, Where AI Flags the Threats and Humans Decide Whether to Call the Police Wiki Serving a 3-Trillion-Parameter Model Takes 59 Wafers, and Cerebras Has Published the Arithmetic Business Micron's Taiwan, China Unions Want 15 Percent of Operating Profit. One Plant Settled, One Walked Out AI Agents MCP Adds Durable Tasks and Mid-Job Questions, Handing Server Authors a Duplicate-Charge Problem News Only 42 Percent of Alpha School Students Hit the 2x Growth Its AI Model Promises, New York Times Finds Opinion AI Made Junior Patent Lawyers Instantly Better. Three Months Later, Only the Seniors Had Kept the Gains News TikTok's US Joint Venture Is Hunting Four Gigawatts of Computing Power by the Early 2030s Business Google Expands Its Virtual Power Plant Tenfold to 1,070 Megawatts, Paying Homes and Businesses to Step Aside for AI Data Centers Security Inside Meta's Law Enforcement Outreach Team, Where AI Flags the Threats and Humans Decide Whether to Call the Police

Only 42 Percent of Alpha School Students Hit the 2x Growth Its AI Model Promises, New York Times Finds

News · By: AIReport Admin · 1 hour ago · 👁 9
Only 42 Percent of Alpha School Students Hit the 2x Growth Its AI Model Promises, New York Times Finds

A New York Times investigation of Alpha School, the teacherless AI-first K-8 chain with about 50 campuses and 1,500 students, found that 42 percent of pupils progressed twice as fast as the national median in both math and reading. Among students who started at or below the 50th percentile in math, only 37 percent hit the 2x mark by winter.

Alpha School sells a simple promise: two to three hours of AI-driven academics in the morning, workshops in the afternoon, and twice the learning of a conventional classroom. A New York Times investigation published on October 23 put that promise against the chain's own test data, and the multiplier did not hold for most students.

According to the Times, 42 percent of Alpha students reached the network's 2x growth standard in both math and reading. The gap was widest in the subject the model leans on hardest. About a fifth of Alpha's K-8 students scored at or below the 50th percentile nationally in math last fall, and only 37 percent of that group met the 2x growth target by winter.

How Alpha arrives at 2x matters as much as the number. The chain divides a student's MAP test growth by the median expected growth and treats the ratio as a multiplier: a student who gains 14 points where 7 were expected is counted as 2x. That is a legitimate growth measure, but it measures pace against a national median rather than the school's own contribution, and it says nothing about students who arrived already ahead.

The operation is small enough to inspect closely: roughly 50 campuses and about 1,500 students, with roots in a single Austin school in 2014. Mornings are software, afternoons are projects, public speaking, sport and vocational workshops, and the adults in the building are positioned as guides who manage motivation rather than deliver instruction.

Retention is the other number the Times surfaced. Some families withdraw after concluding that the model demands an unusual amount of intrinsic motivation and self-discipline from children, which is a quiet way of saying that the format shifts part of the teaching burden onto the household.

None of that has slowed expansion. Alpha is preparing openings in Chicago, Atlanta, Charlotte, Raleigh, Santa Monica, Palo Alto and the East Bay, and its software is already used by about 2,600 public school students across districts in Texas and South Carolina, according to the Times.

The honest read is narrower than either side's talking points. Alpha's students post high scores; the dispute is about attribution and about how many children the format actually fits. A growth ratio computed against a national median cannot answer that, and neither can a marketing page.

What would settle it: cohort baselines published before enrollment, independent administration of the tests, and retention disclosed next to growth. Until a chain does that, twice as fast remains a claim about the median student who stays, which is exactly the student every school already serves well.

Comments (0)

Log in to join the discussion

Log In

No comments yet