Episode Summary
Executive Summary: The episode argues that putting student-facing screens and AI into classrooms has weakened attention, deep reading, and cognitive development. Jared Cooney Horvath says the best evidence favors analog learning, that tech often leads kids to multitask or outsource thinking, and that AI will worsen the problem unless students first build deep knowledge. He urges parents and schools to favor paper, printouts, and opt-outs from non-essential ed tech.
Main Topics: The 'digital delusion' and failed assumptions about ed tech: Horvath says schools adopted a mistaken belief that technology automatically improves learning, but evidence suggests the opposite: more student-facing tech correlates with weaker learning and cognitive skills. How the brain learns best: He explains that attention, memory, and spatial context support learning, which is why writing on paper and reading books often outperform screens, especially for depth and retention. Multitasking and attention economy harms: Laptop and tablet use trains students to multitask, which undermines focused attention and makes it hard for children to stay on-task when the same devices are used for schoolwork. Digital literacy versus tool use: Horvath distinguishes between knowing how to use tech and knowing how to use it meaningfully; he argues schools should build knowledge first, then introduce tools like AI as offloading aids for experts, not outsourcing crutches for novices. Personalized learning and AI critiques: He argues that individualized digital learning often fails for most students because they lack self-regulation, and that AI is volatile, hallucination-prone, and likely to increase inconsistent or incorrect learning. Policy backlash and a return to analog: The conversation highlights growing legislative pushback in the US and Australia, with Horvath calling for analog alternatives, printer use at home, and opt-out rights for families. School’s civic function and social cohesion: He says school should unify students around shared knowledge and debate; hyper-individualized digital pathways can fragment this common ground and weaken democratic cohesion.
Key Arguments: Student-facing ed tech is not neutral; it changes cognition by encouraging multitasking and shallow engagement rather than deep learning. Cognitive scores historically rose with more schooling, but Gen Z is the first generation to see declines despite spending more time in school, suggesting the tools used in classrooms are a major factor. Screens remove the spatial stability of print, which helps memory and deep reading; as a result, students skim rather than read carefully after only a few minutes. Teaching children to use computers or AI does not necessarily improve digital literacy; broad tech instruction rose while measured digital literacy fell. AI is useful only when the user already has expertise to evaluate output; for novices, it becomes outsourcing that bypasses learning entirely. Personalized digital learning works mainly for drill and practice, not for initial learning, because most students lack the self-regulation skills needed to learn independently from a screen. AI introduces inconsistent mistakes and hallucinations, making it a poorer learning environment than vetted digital tools or analog worksheets. School’s role is not just individual optimization but collective formation; hyper-personalized streams reduce shared understanding and civic debate. The most effective near-term response is a return to analog methods at home and in classrooms, plus legal opt-outs for non-essential tech. The harms are not permanent; the brain is plastic, so shifting back to better learning environments can improve outcomes.
Data Points: Time students use computers for multitasking: over 2,500 hours every year - Horvath says kids spend this much time using computers in ways that train multitasking rather than focused learning. Home study before multitasking begins: about 6 minutes - He claims students given a computer at home for homework study on average only six minutes before multitasking. Teacher-present classroom off-task time: 24 to 38 minutes of every hour - He says students on devices in class are off-task or multitasking for much of each lesson. Rise in explicit computer-use teaching: about 600% increase - He cites the ICILS/ISILs digital literacy test and says schools increased direct teaching of computer use dramatically from 2013 to 2023. Digital literacy threshold change: about 11% drop - He says the share of kids reaching the minimum threshold of digital literacy fell over the same period. Students able to learn well on screens: about 5% - He estimates only a small minority can independently learn well via self-regulated screen-based instruction. Students who struggle with screen-based self-learning: about 95% - He argues most children need teacher support and shared classroom instruction rather than individualized digital pathways. States with ed tech bills in the US: 31 states - He says legislative momentum against classroom tech has expanded rapidly by early in the year. States with ed tech bills in January: 3 states - He contrasts early-year momentum with the current number of states considering bills. Digital tutoring efficacy change with AI: worse than non-AI digital tutoring - He says recent studies show digital tutoring systems lose effectiveness when AI is added. AI mistake rate: 20% to 35% of mistakes shift moment by moment - He argues AI output is inconsistent, making errors difficult for teachers or students to detect. Australia NAPLAN literacy concern: over a third of kids lack basic literacy - He cites recent Australian test results as evidence that current schooling outcomes are deteriorating.
Pivotal Quotes: "technology equals learning" — Dr. Jared Cooney Horvath: He describes the core mistaken belief behind the rise of ed tech in schools. "As soon as we inject tech into education, the more learning development and cognitive skills actually start to decrease pretty significantly." — Dr. Jared Cooney Horvath: He summarizes his thesis that classroom technology undermines rather than improves learning. "offloading versus outsourcing" — Dr. Jared Cooney Horvath: He uses this distinction to explain why experts can use tools productively while children often cannot.
Implications: The episode urges schools and parents to slow ed-tech adoption, prioritize paper-based learning, and treat AI as an expert tool rather than a student shortcut. It suggests policy will increasingly favor opt-outs, analog alternatives, and tighter limits on classroom screens.