Recent testimony to a senate committee by Dr. Jared Horvath, neuroscientist and educator, presented studies that showed alarming downward trends in student cognitive development since the mid-2000s across developed nations. This was the first decline since the beginning of the 20th century, when standardized education sparked a century long improvement in those same cognitive abilities.
Dr. Horvath concludes that the aggressive introduction of technology and extended periods of screen time have had a deleterious effect on education and cognitive development. This testimony will no doubt be strongly contested by the technology establishment, as it calls into question one of the key claims that AI makes for its future profitability – the goal of enhancing human intelligence and lives across class and cultural barriers.
I have some anecdotal evidence that supports Horvath’s thesis as I have spent the last 6 months tutoring Hispanic high school kids in math. All of their work is on Chromebooks. There is no math textbook and every problem is presented and worked on the computer. The students enter every calculation into a subject line as part of their solution and use associated calculators to perform them. Not a single student has immediate recall of multiplication tables (8 times 7 for example) or knows how to add fractions using common denominators. Even basic addition is wanting.
This would perhaps be ok if their command of higher-level concepts were solid, but their work is almost totally mimicry or use of AI in other web windows. They struggle to do any problems that truly require deeper understanding of the theory. These students are much less capable than the very weakest students I taught in my three years as a high school mathematics teacher from 2006-2009, before any significant introduction of technology.
Schools and universities are desperate to avoid the stigma of missing the boat on AI, so they appear to be embracing AI and other technology components with willful abandon and probably spending funds that could have been used to hire and train teachers. The hype and pressure from AI companies and their legions of pundits is immense.
Horvath’s research on cognitive development theorizes that human evolution is much better adapted to person-to-person learning focusing on a single task. The combination of screen time trances, application multi-tasking and non-academic app distraction create a net negative effect when technology is the centerpiece of instruction. His survey of research indicates that reading comprehension and retention are much stronger on paper, as is note-taking, particularly for complex or extended texts or tasks.
Technology used for practice and remediation to support learning objectives has been shown to be net positive in many studies, but this type of limited utilization is anathema to the tech titans who see their products being center stage and ubiquitous. They will no doubt be relentless in their insistence that the future is in AI education and that it is only a matter of educators adapting to the new tools.
Public policy failed us completely during the advent of social media and legislators allowed tech apps to foul the waters of social intercourse and adolescent mental health in the name of so-called free speech and the free market. Let us hope that Horvath’s warnings will be heeded and that the avalanche of AI-based education ‘miracles’ will be subjected to careful scrutiny, monitoring and control before we find that we have irreversibly damaged our species. But then again, those future generations will be so much more easily manipulated by AI, so perhaps it is all just part of the master plan!