Tom Werner/Getty Images
The tech industry is ensuring that kids are hooked to AI for generations – by instilling their tendencies deep into the education system, long before they understand the impact of technology on young minds.
Top leaders in the field, from Microsoft to OpenAI, are pouring millions of dollars into schools, colleges, and universities, often providing students access to their AI products. justification, stated in a fresh new York Times Piece Both the tech companies and the teachers receiving funding believe these tools will accelerate learning and prepare students for a world driven by AI.
But the truth beyond this propaganda is much murkier and darker. While some research suggests that AI actually disrupts learning, a notable study conducted by researchers at Microsoft and Carnegie Mellon showed that it weakens critical thinking skills.
Even more immediately, the safety of AI chatbots is looking more questionable by the day, as significant media and clinical attention is being given to the phenomenon of so-called AI psychosis, in which users – many of them teens and young adults – go into delusional psychotic spirals through interactions with human-sounding AI. Some of these spirals have even reached the point of suicide and murder.
New technology always causes conflict in education, and teachers once became aware of calculators as well. But never before had a device outsourced the task of cognition so thoroughly – Nevermind acted as a personal assistant, Friend, or lover.
Worst of all, AI companies are rapidly entering education before the dust has settled on any of these important questions. In the US, Miami-Dade County Public Schools, the country’s third-largest school system, has deployed a version of Google’s Gemini chatbot to its more than 100,000 high school students. NYT noted. On the other side of the learning spectrum, Microsoft, OpenAI and Anthropic have invested more than $23 million in one of the country’s largest teacher unions to provide training to members to use their AI products.
Abroad, Elon Musk’s AI company, xAI, announced It deployed its chatbot Grok to more than 5,000 public schools in El Salvador last month, calling it “the world’s first nationwide AI-powered education program.” Last June, Microsoft partnered with the Ministry of Education in Thailand to provide free online lessons on the use of AI to hundreds of thousands of students. NYTBefore later announcing free AI training for nearly the same number of teachers.
Some experts fear we are making the same mistake that happened with a global effort to increase access to computers, called one laptop per child, which did not improve students’ scores or their cognitive abilities. studies quoted by Times,
“The side effects of one laptop per child include wasteful spending and poor learning outcomes,” Steven Vosloo, a digital policy expert at UNICEF, wrote in a recently viewed post. NYT“Unchecked use of AI systems can actively deskilled students and teachers,”
One argument being made is that placing children in a controlled environment such as school can better prepare them for their inevitable encounters with AI chatbots, giving them the knowledge to use them safely and effectively. Yet the top AI companies bankrolling billions have shown that they are still unable to keep a tight grip on their tools and ensure they remain consistently secure. (OpenAI recently acknowledged that its own data showed that perhaps half a million ChatGPT users were having conversations that showed signs of psychosis, a revelation that hasn’t stopped it from powering its big language models to children’s toys.)
We’re just beginning to grapple with the consequences that another digital innovation, social media, has had on children and teens — and now the tech industry wants to take part in the next digital experiment before we even know if it’s safe.
The truth, frankly, is that AI companies don’t know whether their products are safe or useful for students. But in the race to gain market share in a competitive space, they’re wasting no time in finding out.
More on AI: New parents mocked for letting ChatGPIT name their baby
