How AI Impacts Memory Function

Your brain on AI: One of the first studies measuring what ChatGPT use does to our brain MIT researchers tracked 54 people writing essays using ChatGPT, web search, or just their brains—while monitoring neural activity with EEG. The findings are striking: 🧠 Brain connectivity weakened with more AI support. ChatGPT users showed the least neural engagement. 🔍 Memory collapsed. 83% of ChatGPT users couldn't quote their own essays minutes later, vs. near-perfect recall without AI. ⚡ "Cognitive debt" accumulated. When ChatGPT users later wrote without AI, their brains showed weakened connectivity compared to those who practiced unassisted writing. 🎨 Creativity declined. AI-assisted essays were statistically more uniform and less original. The twist: Strategic timing matters. Using AI after initial self-driven effort preserved better cognitive engagement than consistent AI use from the start. This isn't anti-AI—it's about understanding the trade-offs. While AI-generated essays scored well initially, participants showed signs of cognitive atrophy: diminished critical thinking, reduced memory encoding, and less ownership of their work. The takeaway: We need to enhance, not replace, human thinking as we integrate these powerful tools. Full study here: https://lnkd.in/e-6urMD8 Note: This is a pre-print study awaiting peer review.

🧠 Your Brain Is Quietly Paying a Price for Using ChatGPT We spend hours with LLMs like ChatGPT. But are we fully aware of what they’re doing to our brains? A new study from MIT delivers a clear message: The more we rely on AI to generate and structure our thoughts, the more we risk losing touch with essential cognitive processes — creativity, memory, and critical reasoning. 📊 Key insight? When students wrote essays using GPT-4o, real-time EEG data showed a significant decline in activity across brain regions tied to executive control, semantic processing, and idea generation. When those same students later had to write without AI assistance, their performance didn’t just drop — it collapsed. 🔬 What they did: 54 students wrote SAT-style essays across multiple sessions, while high-density EEG tracked information flow between 32 brain regions. Participants were split across three tools: → Solo writing (“Brain-only”) → Google Search → GPT-4o (LLM-assisted) In the final round, the groups switched: GPT users wrote unaided, and unaided writers used GPT. (LLM→Brain and Brain→LLM) ⚡ What they found: Neural dampening: Full reliance on the LLM led to the weakest fronto-parietal and temporal connectivity — signaling lighter executive function and shallower semantic engagement. Sequence effects: Writers who began solo and then layered on GPT showed increased brain-wide activity — a sign of active cognitive engagement. The reverse group (starting with GPT) showed the lowest coordination and overused LLM-preferred vocabulary. Memory failures: In their very first AI-assisted session, no GPT users could recall a single sentence they had just written — while most solo writers could. Cognitive debt: Repeated LLM use led to narrower idea generation and reduced topic diversity — making recovery without AI more difficult. 🌱 What does this mean for us? LLMs make content creation feel frictionless. But that very convenience comes at a cost: Diminished engagement. Lower memory. Narrower thinking. If we want to preserve intellectual independence and the ability to truly think, we need to use LLMs with intention. →Use them too soon, and the brain goes quiet. →Use them after thinking independently — and they amplify our output. ✨ Hybrid workflows are the way forward: Start with your own cognition, then apply LLMs to sharpen, not replace. The most irreplaceable kind of AI will always be Actual Intelligence. 👉 Full study (with TL;DR + summary table): https://zurl.co/0hnox

Why the Generative AI Industry Has Launched a Coordinated Attack on the MIT Brain Study It threatens their $100 billion myth A new study from MIT Media Lab⁽¹⁾ just did what few have dared: it questioned the cognitive cost of relying on Generative AI too soon. The study observed students writing SAT-style essays in three conditions: No tools (Brain-only) Google-style search ChatGPT (LLM) The results? 83% of ChatGPT users could not recall their own sentence minutes after writing it. Only 11% of the control group had this issue. EEG scans showed weaker neural activity in those using ChatGPT, especially in alpha and beta frequency bands linked to memory, attention, and integration. Essays written with ChatGPT were more formulaic and less original, and users reported less sense of ownership over their writing. Using ChatGPT didn’t just change how students wrote. It changed how they remembered. It changed how they thought. So why the backlash? Because these findings strike at the heart of two foundational industry myths: Lie #1: “AI makes us smarter.” What the MIT study suggests is that if you use it too early, AI may short-circuit the parts of the brain responsible for learning, retention, and creativity. Lie #2: “Generative AI is universally empowering.” Maybe, but only when it augments us. Not when it replaces our first thoughts. What followed was a wave of pushback from AI insiders and enthusiasts: “The sample size is too small.” “EEG doesn’t prove disengagement.” “It’s alarmist.” “But what about accessibility?” In all fairness, the study was far from perfect, though these issues do not discredit the findings. Here are some of the concerns one might credibly raise: Sample size (n=54) is modest EEG signals are correlational, not causal Still, none of these critiques refute the core pattern: The more you outsource early thinking, the less engaged your mind becomes. Users who started with AI continued to show less neural activation even when later asked to write on their own. The real story isn’t that MIT exposed some fatal flaw in AI. The real story is how fast the industry tried to bury it. ******************************************************************************** The trick with technology is to avoid spreading darkness at the speed of light. I’m the Founder & CEO of Curiouser.AI, a Generative AI platform and strategic advisory focused on elevating organizations and augmenting human intelligence through strategic coaching and values-based leadership. I also teach Marketing and AI Ethics at UC Berkeley. If you're a CEO or board member committed to building a stronger, values-driven organization in the age of AI, reach out. Visit curiouser.ai, DM me, or connect on Hubble: https://lnkd.in/gphSPv_e Sources: Kosmyna, N. et al. Your Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant for Essay Writing Task. MIT Media Lab, June 2025. Ibid. Recall data and EEG analysis referenced from preprint version.

𝐍𝐨, 𝐲𝐨𝐮𝐫 𝐛𝐫𝐚𝐢𝐧 𝐝𝐨𝐞𝐬 𝐧𝐨𝐭 𝐩𝐞𝐫𝐟𝐨𝐫𝐦 𝐛𝐞𝐭𝐭𝐞𝐫 𝐚𝐟𝐭𝐞𝐫 𝐋𝐋𝐌 𝐨𝐫 𝐝𝐮𝐫𝐢𝐧𝐠 𝐋𝐋𝐌 𝐮𝐬𝐞. See our paper for more results: "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task" (link in the comments). For 4 months, 54 students were divided into three groups: ChatGPT, Google -ai, and Brain-only. Across 3 sessions, each wrote essays on SAT prompts. In an optional 4th session, participants switched: LLM users used no tools (LLM-to-Brain), and Brain-only group used ChatGPT (Brain-to-LLM). 👇 𝐈. 𝐍𝐋𝐏 𝐚𝐧𝐝 𝐄𝐬𝐬𝐚𝐲 𝐂𝐨𝐧𝐭𝐞𝐧𝐭 - LLM Group: Essays were highly homogeneous within each topic, showing little variation. Participants often relied on the same expressions or ideas. - Brain-only Group: Diverse and varied approaches across participants and topics. - Search Engine Group: Essays were shaped by search engine-optimized content; their ontology overlapped with the LLM group but not with the Brain-only group. 𝐈𝐈. 𝐄𝐬𝐬𝐚𝐲 𝐒𝐜𝐨𝐫𝐢𝐧𝐠 (𝐓𝐞𝐚𝐜𝐡𝐞𝐫𝐬 𝐯𝐬. 𝐀𝐈 𝐉𝐮𝐝𝐠𝐞) - Teachers detected patterns typical of AI-generated content and scoring LLM essays lower for originality and structure. - AI Judge gave consistently higher scores to LLM essays, missing human-recognized stylistic traits. 𝐈𝐈𝐈: 𝐄𝐄𝐆 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 Connectivity: Brain-only group showed the highest neural connectivity, especially in alpha, theta, and delta bands. LLM users had the weakest connectivity, up to 55% lower in low-frequency networks. Search Engine group showed high visual cortex engagement, aligned with web-based information gathering. 𝑺𝒆𝒔𝒔𝒊𝒐𝒏 4 𝑹𝒆𝒔𝒖𝒍𝒕𝒔: - LLM-to-Brain (🤖🤖🤖🧠) participants underperformed cognitively with reduced alpha/beta activity and poor content recall. - Brain-to-LLM (🧠🧠🧠🤖) participants showed strong re-engagement, better memory recall, and efficient tool use. LLM-to-Brain participants had potential limitations in achieving robust neural synchronization essential for complex cognitive tasks. Results for Brain-to-LLM participants suggest that strategic timing of AI tool introduction following initial self-driven effort may enhance engagement and neural integration. 𝐈𝐕. 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫𝐚𝐥 𝐚𝐧𝐝 𝐂𝐨𝐠𝐧𝐢𝐭𝐢𝐯𝐞 𝐄𝐧𝐠𝐚𝐠𝐞𝐦𝐞𝐧𝐭 - Quoting Ability: LLM users failed to quote accurately, while Brain-only participants showed robust recall and quoting skills. - Ownership: Brain-only group claimed full ownership of their work; LLM users expressed either no ownership or partial ownership. - Critical Thinking: Brain-only participants cared more about 𝘸𝘩𝘢𝘵 and 𝘸𝘩𝘺 they wrote; LLM users focused on 𝘩𝘰𝘸. - Cognitive Debt: Repeated LLM use led to shallow content repetition and reduced critical engagement. This suggests a buildup of "cognitive debt", deferring mental effort at the cost of long-term cognitive depth. Support and share! ❤️ #MIT #AI #Brain #Neuroscience #CognitiveDebt

If you know me, you know I have always been a bit leery of AI. I view it as a tool, and a useful one, but one that too many people go overboard with. I have always been most afraid that it was going to result in people becoming too isolated. I never anticipated it was going to make people.... less intelligent. I'm not the one suggesting it. None other than MIT is. In order to examine the 'cognitive costs' of relying on LLM's, scientists from MIT's Media Lab tracked 54 students over a 4 month period. The results? The group who used LLM's such as ChatGPT to write essays "showed poorer memory, reduced brain activity and weaker engagement than those who used other methods." Additionally, the LLM group's participants "performed worse than their counterparts in the Brain-only group at all levels: neural, linguistic, scoring." The LLM users engaged less over each session, began to rely on ChatGPT more and more as the study went on, and ultimately felt less ownership of their work (which was also judged to be less unique). They often failed to accurately quote from their own work, suggesting an impact on memory and data retention. And here's the real worry. When the brain-only and LLM using groups of students traded places, the group that had previously relied on AI to do their writing did not return to their pre-testing levels of brain activity. If you're outsourcing all of your writing to ChatGPT, what are you sacrificing? Thank you to Henna Pryor, CSP, I first read about this study on your Instagram and have gone down a rabbit hole on the subject. ✳️ The t-shirts in the photo? A gift from my brilliant friend Evan Benjamin - because when he asked AI to write "Trustworthy" in an image, he got "Trustwobbly" instead. It seemed the perfect photo to go with this post.

🏁 The Sunday Curve | Week 6 🧠 MIT Media Lab study finds LLM use dampens brain engagement A recent preprint study by MIT’s Media Lab ("Your Brain on ChatGPT: Accumulation of Cognitive Debt...") followed 54 young adults across four months, comparing three groups writing essays: 🔘 Brain-only: no external help 🔘 Search engine (like Google) 🔘 LLM‑assisted (ChatGPT)  Study https://lnkd.in/g93WJSuT Key findings: 🔘 EEG scans showed the LLM group had the weakest neural connectivity, especially in alpha and beta bands markers of executive function, attention, and memory. 🔘 LLM writers produced essays that were more uniform, less original, and often lacked personal ownership. Over 80% couldn’t recall their own content properly https://lnkd.in/g6_u5b6u 🔘 Even after switching away from AI, LLM users showed persistently lower brain activity, while brain-only participants who adopted AI later performed better. 🔘 The researchers describe this as "cognitive debt" short-term speed at the cost of deeper thinking, memory, creativity, and engagement  🔘 Experts caution against blanket conclusions; impact depends on how LLMs are used. When used as tutors (especially in educational settings), they can boost learning https://lnkd.in/grDGkGZk 🔘 The Times, Economic Times, NY Post, Time, and Times of India highlight diminished brain activity with LLM use and emphasize the need for mindful integration https://lnkd.in/ggRWRAWC 🔘 A Wall Street Journal op‑ed reflects on individuals recognizing AI was weakening their cognitive agility, and urges balanced habits https://lnkd.in/gQRcU2i7 ✅ Finally…. Using LLMs for shortcuts—such as having ChatGPT write essays can measurably reduce brain activity, memory retention, originality, and engagement.  🔘 Look for guidelines on balanced AI integration, especially in education and workplace settings. 🔘 Researchers are also exploring how LLMs influence skills beyond writing, like programming. ✅ Watch for my take on this next week. Photo credit: #AI #ArtificialIntelligence #Chatgpt #LLM #TheSundayCurve

As a research leader at LinkedIn and psychology faculty member at USC, I’m deeply interested in how emerging technologies like AI are reshaping how we think, learn, and create — not just behaviorally, but cognitively and systemically. A recent TIME article (https://lnkd.in/gP89SwPv) covers an early-release MIT study that used EEG to measure brain activity as participants wrote essays using: 🧠 Just their brain 🔍 Google Search 🤖 ChatGPT The early findings: -ChatGPT users showed the weakest neural connectivity, especially in alpha, theta, and delta bands — linked to creativity, memory, and semantic processing -Their writing became more predictable and less original -Even after switching tools, cognitive engagement remained low, suggesting lingering offloading effects The study hasn’t been peer-reviewed yet, and like all EEG research, its insights are correlational — not causal. EEG data reflects surface-level patterns of brain activity, and interpretation requires caution. Still, the results are compelling enough that the authors chose to release the paper early due to the potential implications at scale. As AI becomes more deeply embedded in how we learn and work, studies like this serve as a critical reminder: We must design for engagement, resilience, and trust — not just efficiency and output. This is our moment to guide AI’s role in education and work — ensuring it strengthens critical thinking, equity, and meaningful human connection, rather than unintentionally diminishing them. #UXResearch #CognitivePsychology #Neuroscience #AI #FutureOfLearning #HumanCenteredAI #PsychologicalSafety #EquityInTech #EdTech