{"id":133589,"date":"2025-09-10T20:18:08","date_gmt":"2025-09-10T20:18:08","guid":{"rendered":"https:\/\/www.newsbeep.com\/au\/133589\/"},"modified":"2025-09-10T20:18:08","modified_gmt":"2025-09-10T20:18:08","slug":"how-does-ai-affect-how-we-learn-a-cognitive-psychologist-explains-why-you-learn-when-the-work-is-hard","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/au\/133589\/","title":{"rendered":"How does AI affect how we learn? A cognitive psychologist explains why you learn when the work is hard"},"content":{"rendered":"<p>When OpenAI released \u201c<a href=\"https:\/\/openai.com\/index\/chatgpt-study-mode\/\" rel=\"nofollow noopener\" target=\"_blank\">study mode<\/a>\u201d in July 2025, the company touted ChatGPT\u2019s educational benefits. \u201cWhen ChatGPT is prompted to teach or tutor, it can significantly improve academic performance,\u201d <a href=\"https:\/\/venturebeat.com\/ai\/chatgpt-just-got-smarter-openais-study-mode-helps-students-learn-step-by-step\" rel=\"nofollow noopener\" target=\"_blank\">the company\u2019s vice president of education told reporters<\/a> at the product\u2019s launch. But any dedicated teacher would be right to wonder: Is this just marketing, or does scholarly research really support such claims?<\/p>\n<p>While generative AI tools are moving into classrooms at lightning speed, robust research on the question at hand hasn\u2019t moved nearly as fast. Some early studies have shown benefits for certain groups such as <a href=\"https:\/\/doi.org\/10.1016\/j.caeai.2023.100147\" rel=\"nofollow noopener\" target=\"_blank\">computer programming students<\/a> and <a href=\"https:\/\/doi.org\/10.1186\/s41239-023-00425-2\" rel=\"nofollow noopener\" target=\"_blank\">English language learners<\/a>. And there have been a number of other optimistic studies on AI in education, such as one <a href=\"https:\/\/doi.org\/10.1057\/s41599-025-04787-y\" rel=\"nofollow noopener\" target=\"_blank\">published in the journal Nature<\/a> in May 2025 suggesting that chatbots may aid learning and higher-order thinking. But scholars in the field have pointed to <a href=\"https:\/\/buildcognitiveresonance.substack.com\/p\/something-rotten-in-ai-research\" rel=\"nofollow noopener\" target=\"_blank\">significant methodological weaknesses<\/a> in <a href=\"https:\/\/wesstrabelsi.substack.com\/p\/the-good-the-bad-and-the-ugly-science\" rel=\"nofollow noopener\" target=\"_blank\">many of these research papers<\/a>.<\/p>\n<p>Other studies have painted a grimmer picture, suggesting that AI may impair performance or cognitive abilities such as <a href=\"https:\/\/doi.org\/10.3390\/soc15010006\" rel=\"nofollow noopener\" target=\"_blank\">critical thinking skills<\/a>. <a href=\"https:\/\/doi.org\/10.3390\/app14104115\" rel=\"nofollow noopener\" target=\"_blank\">One paper showed<\/a> that the more a student used ChatGPT while learning, the worse they did later on similar tasks when ChatGPT wasn\u2019t available.<\/p>\n<p>In other words, early research is only beginning to scratch the surface of how this technology will truly affect learning and cognition in the long run. Where else can we look for clues? As a cognitive psychologist who has <a href=\"https:\/\/doi.org\/10.1177\/00986283241305398\" rel=\"nofollow noopener\" target=\"_blank\">studied how college students are using AI<\/a>, I have found that my field offers valuable guidance for identifying when AI can be a brain booster and when it risks becoming a brain drain.<\/p>\n<p>Skill comes from effort<\/p>\n<p>Cognitive psychologists have argued that our thoughts and decisions are the result of two processing modes, commonly denoted as <a href=\"https:\/\/doi.org\/10.1017\/S0140525X00003435\" rel=\"nofollow noopener\" target=\"_blank\">System 1 and System 2<\/a>. <\/p>\n<p>The former is a system of pattern matching, intuition and habit. It is fast and automatic, requiring little conscious attention or cognitive effort. Many of our routine daily activities \u2013 getting dressed, making coffee and riding a bike to work or school \u2013 fall into this category. System 2, on the other hand, is generally slow and deliberate, requiring more conscious attention and <a href=\"https:\/\/doi.org\/10.1037\/bul0000443\" rel=\"nofollow noopener\" target=\"_blank\">sometimes painful cognitive effort<\/a>, but often yields more robust outputs.<\/p>\n<p>We need both of these systems, but gaining knowledge and mastering new skills depend heavily on System 2. Struggle, <a href=\"https:\/\/doi.org\/10.1111\/cogs.12107\" rel=\"nofollow noopener\" target=\"_blank\">friction<\/a> and <a href=\"https:\/\/doi.org\/10.1146\/annurev-neuro-072116-031526\" rel=\"nofollow noopener\" target=\"_blank\">mental effort<\/a> are crucial to the cognitive work of <a href=\"https:\/\/doi.org\/10.1037\/a0037559\" rel=\"nofollow noopener\" target=\"_blank\">learning<\/a>, <a href=\"https:\/\/doi.org\/10.3389\/fpsyg.2015.01054\" rel=\"nofollow noopener\" target=\"_blank\">remembering<\/a> and <a href=\"https:\/\/doi.org\/10.1016\/j.neuron.2015.12.029\" rel=\"nofollow noopener\" target=\"_blank\">strengthening connections in the brain<\/a>. Every time a confident cyclist gets on a bike, they rely on the hard-won pattern recognition in their System 1 that they previously built up through many hours of effortful System 2 work spent learning to ride. You don\u2019t get mastery and you can\u2019t <a href=\"https:\/\/doi.org\/10.1037\/xlm0000578\" rel=\"nofollow noopener\" target=\"_blank\">chunk information efficiently<\/a> for higher-level processing without first putting in the cognitive effort and strain.<\/p>\n<p>I tell my students the brain is a lot like a muscle: It takes genuine <a href=\"https:\/\/bjorklab.psych.ucla.edu\/wp-content\/uploads\/sites\/13\/2016\/04\/EBjork_RBjork_2011.pdf\" rel=\"nofollow noopener\" target=\"_blank\">hard work to see gains<\/a>. Without challenging that muscle, it won\u2019t grow bigger.<\/p>\n<p>What if a machine does the work for you?<\/p>\n<p>Now imagine a robot that accompanies you to the gym and lifts the weights for you, no strain needed on your part. Before long, your own muscles will have atrophied and you\u2019ll become reliant on the robot at home even for simple tasks like moving a heavy box. <\/p>\n<p>AI, used poorly \u2013 to complete a quiz or write an essay, say \u2013 lets students bypass the very thing they need to develop knowledge and skills. It takes away the mental workout.<\/p>\n<p>Using technology to effectively offload cognitive workouts can have a detrimental effect on <a href=\"https:\/\/doi.org\/10.1177\/17470218211008060\" rel=\"nofollow noopener\" target=\"_blank\">learning and memory<\/a> and can cause people to misread their own understanding or abilities, leading to what psychologists call <a href=\"https:\/\/doi.org\/10.1016\/j.tics.2016.07.002\" rel=\"nofollow noopener\" target=\"_blank\">metacognitive errors<\/a>. Research has shown that habitually offloading car navigation to GPS may <a href=\"https:\/\/doi.org\/10.1038\/s41598-020-62877-0\" rel=\"nofollow noopener\" target=\"_blank\">impair spatial memory<\/a> and that using an external source like Google to answer questions <a href=\"https:\/\/doi.org\/10.1073\/pnas.2105061118\" rel=\"nofollow noopener\" target=\"_blank\">makes people overconfident<\/a> in their own personal knowledge and memory. <\/p>\n<p>            <img decoding=\"async\" alt=\"Girl doing school with phone and notebook.\" class=\"lazyload\" src=\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2025\/09\/file-20250909-56-4nxavu.jpg\"  \/><\/p>\n<p>              Learning and mastery come from effort, whether that\u2019s done with a powerful chatbot or AI tutor or not, but educators and students need to resist outsourcing that work.<br \/>\n              <a class=\"source\" href=\"https:\/\/www.gettyimages.com\/detail\/photo\/young-girl-uses-ai-virtual-assistant-to-do-royalty-free-image\/2204906139\" rel=\"nofollow noopener\" target=\"_blank\">Francesco Carta fotografo via Getty Images<\/a><\/p>\n<p>Are there similar risks when students hand off cognitive tasks to AI? <a href=\"https:\/\/doi.org\/10.1016\/j.chb.2024.108386\" rel=\"nofollow noopener\" target=\"_blank\">One study found<\/a> that students researching a topic using ChatGPT instead of a traditional web search had lower cognitive load during the task \u2013 they didn\u2019t have to think as hard \u2013 and produced worse reasoning about the topic they had researched. Surface-level use of AI may mean less cognitive burden in the moment, but this is akin to letting a robot do your gym workout for you. It ultimately leads to poorer thinking skills.<\/p>\n<p>In another study, students <a href=\"https:\/\/doi.org\/10.1111\/bjet.13544\" rel=\"nofollow noopener\" target=\"_blank\">using AI to revise their essays<\/a> scored higher than those revising without AI, often by simply copying and pasting sentences from ChatGPT. But these students showed no more actual knowledge gain or knowledge transfer than their peers who worked without it. The AI group also engaged in fewer rigorous System 2 thinking processes. The authors warn that such \u201cmetacognitive laziness\u201d may prompt short-term performance improvements but also lead to the stagnation of long-term skills.<\/p>\n<p>Offloading can be useful once foundations are in place. But those foundations can\u2019t be formed unless your brain does the initial work necessary to encode, connect and understand the issues you\u2019re trying to master.<\/p>\n<p>Using AI to support learning<\/p>\n<p>Returning to the gym metaphor, it may be useful for students to think of AI as a personal trainer who can keep them on task by <a href=\"https:\/\/doi.org\/10.1016\/j.neuroimage.2018.12.003\" rel=\"nofollow noopener\" target=\"_blank\">tracking and scaffolding<\/a> learning and <a href=\"https:\/\/doi.org\/10.1073\/pnas.2111785119\" rel=\"nofollow noopener\" target=\"_blank\">pushing them<\/a> to work harder. AI has great potential as a <a href=\"https:\/\/doi.org\/10.1371\/journal.pone.0304013\" rel=\"nofollow noopener\" target=\"_blank\">scalable learning tool<\/a>, <a href=\"https:\/\/www.educationnext.org\/ai-tutors-hype-or-hope-for-education-forum\/\" rel=\"nofollow noopener\" target=\"_blank\">an individualized tutor<\/a> with a vast knowledge base that never sleeps. <\/p>\n<p>AI technology companies are seeking to design just that: the ultimate tutor. In addition to <a href=\"https:\/\/www.insidehighered.com\/news\/tech-innovation\/artificial-intelligence\/2025\/08\/07\/understanding-value-learning-fuels-chatgpts\" rel=\"nofollow noopener\" target=\"_blank\">OpenAI\u2019s entry into education<\/a>, in April 2025 Anthropic released its <a href=\"https:\/\/www.anthropic.com\/news\/introducing-claude-for-education\" rel=\"nofollow noopener\" target=\"_blank\">learning mode<\/a> for Claude. These models are supposed to engage in Socratic dialogue, to pose questions and provide hints, rather than just giving the answers. <\/p>\n<p>Early research indicates <a href=\"http:\/\/dx.doi.org\/10.2139\/ssrn.4895486\" rel=\"nofollow noopener\" target=\"_blank\">AI tutors can be beneficial but introduce problems as well<\/a>. For example, one study found high school students reviewing math with ChatGPT performed worse than students who didn\u2019t use AI. Some students used the base version and others a customized tutor version that gave hints without revealing answers. When students took an exam later without AI access, those who\u2019d used base ChatGPT did much worse than a group who\u2019d studied without AI, yet they didn\u2019t realize their performance was worse. Those who\u2019d studied with the tutor bot did no better than students who\u2019d reviewed without AI, but they mistakenly thought they had done better. So AI didn\u2019t help, and it introduced metacognitive errors. <\/p>\n<p>Even as tutor modes are refined and improved, students have to actively select that mode and, for now, also have to <a href=\"https:\/\/www.insidehighered.com\/news\/tech-innovation\/artificial-intelligence\/2025\/08\/07\/understanding-value-learning-fuels-chatgpts\" rel=\"nofollow noopener\" target=\"_blank\">play along<\/a>, deftly providing context and guiding the chatbot away from worthless, low-level questions <a href=\"https:\/\/doi.org\/10.1007\/978-3-031-92611-2_5\" rel=\"nofollow noopener\" target=\"_blank\">or sycophancy<\/a>. <\/p>\n<p>The latter issues may be fixed with <a href=\"https:\/\/doi.org\/10.48550\/arXiv.2409.01658\" rel=\"nofollow noopener\" target=\"_blank\">better design, system prompts<\/a> and custom interfaces. But the temptation of using default-mode AI to avoid hard work will continue to be a more fundamental and classic problem of teaching, course design and motivating students to avoid shortcuts that undermine their cognitive workout.<\/p>\n<p>As with other complex technologies such as <a href=\"https:\/\/time.com\/6958809\/smartphones-screen-time-danger-pete-etchells\/\" rel=\"nofollow noopener\" target=\"_blank\">smartphones<\/a>, <a href=\"https:\/\/doi.org\/10.1017\/mem.2021.3\" rel=\"nofollow noopener\" target=\"_blank\">the internet<\/a> or even <a href=\"https:\/\/www.perseus.tufts.edu\/hopper\/text?doc=Perseus%3Atext%3A1999.01.0174%3Atext%3DPhaedrus%3Apage%3D274\" rel=\"nofollow noopener\" target=\"_blank\">writing itself<\/a>, it will take more time for researchers to fully understand the true range of AI\u2019s effects on cognition and learning. In the end, the picture will likely be a nuanced one that depends heavily on context and use case. <\/p>\n<p>But what we know about learning tells us that deep knowledge and mastery of a skill will always require a genuine cognitive workout \u2013 with or without AI.<\/p>\n","protected":false},"excerpt":{"rendered":"When OpenAI released \u201cstudy mode\u201d in July 2025, the company touted ChatGPT\u2019s educational benefits. \u201cWhen ChatGPT is prompted&hellip;\n","protected":false},"author":2,"featured_media":133590,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[256,254,255,64,63,105],"class_list":{"0":"post-133589","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-artificial-intelligence","8":"tag-ai","9":"tag-artificial-intelligence","10":"tag-artificialintelligence","11":"tag-au","12":"tag-australia","13":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/133589","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/comments?post=133589"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/posts\/133589\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media\/133590"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/media?parent=133589"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/categories?post=133589"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/au\/wp-json\/wp\/v2\/tags?post=133589"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}