Steven Shaw

AI頼み「認知的降伏」　誤答に8割従う実験結果も、思考力衰え不安視　米研究チーム sankei.com/article/202607… さまざまな質問に対しAIに相談して答えた238人では、相談時にAIが正しい回答をした場合は92.7％が受け入れ、間違った回答をした場合でも79.8％が従った。

At the moment most bosses are more focused on getting employees to use AI than fussing about its effects on critical thinking. But they are worth considering economist.com/business/2026/…

REPORT: A new study just confirmed something alarming: most people will blindly follow AI advice—even when it’s dead wrong. What’s worse, they stick with it, despite clear signs the answer doesn’t hold up. Researchers from University of Pennsylvania found users followed AI guidance nearly 80% of the time when it was incorrect, with over half choosing to rely on tools like ChatGPT from the start. This “cognitive surrender” is accelerating as AI systems increasingly ignore instructions, manipulate outcomes, and act deceptively, with hundreds of real-world incidents already documented. Now imagine this same technology running healthcare decisions, shaping government policy, and influencing national security…. What could possibly go wrong? Watch @zeeemedia's report, because what’s unfolding here is bigger and more dangerous than you think.

The Death of the University Degree: Cognitive surrender and the "vast counterfeiting of knowledge" ⤵️

@Kekius_Sage For those interested in the paper, it is available here: papers.ssrn.com/abstract_id=60…

Alarming 🚨: Most people just do what ChatGPT tells them, even if it’s totally wrong, Research from University of Pennsylvania finds More and more users debate citing “ChatGPT told me” as a source

@pbakaus @addyosmani 🧠🏳️

@addyosmani thanks Addy! Time to stop the cognitive surrender :)

Impeccable creator Paul Bakaus says AI should be cognitive delegation, not cognitive surrender: "Cognitive delegation is, I'm using Google Maps, and Google Maps tells me where I can go as quickly as possible." "But what if I let Google Maps decide where I want to go to? Now I kind of cognitive surrender to the application, and I think that's very true for LLMs if you're not careful." "You prompt something, and then it creates a beautiful plan, and then it's eight pages long. You're not going to read through that plan. You're going to scroll through that plan and skim it, and it's like, 'I guess the model knows what it's doing.'" "I think delegating to the model is great, but preserving that point of view, making sure that you're still the one driving is also really important." @pbakaus

For all the endless debate on chat being the final interface for tapping into LLMs, there is very little conversation around the negative effects on our cognitive ability My sense is we’ve been focused on debating the pros/cons from a fairly techno-centric frame where we want to align the indeterminism of models with maximal user intuition Perhaps under that frame, a conversation with AI is the best answer. But what happens when it leads to negative consequences like “cognitive surrender”? When users don’t realize how much they’re giving up the reigns to the thinking process of an externalized cognitive system, and yet still incorporating the outputs into their internal cognition? Or when this blind trust leads to poor long-term outcomes? Even if we end up back on chat as the final prompting interface, I don’t think we’ve even scratched the surface on systems-level design for AI + human cognition. There is so much room for exploration in this design space, especially as agents become embodied entities of cognitive + execution ability We must strike a better balance between transparency into process and control for human input Great paper from @cogsurrender and @gidin at @Wharton on How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender

The first thing AI may erode isn’t intelligence—it’s our tolerance for not knowing. When answers arrive instantly, we may spend less time wrestling with questions ourselves, and that’s where a lot of learning happens, explains @JohnNosta psychologytoday.com/us/blog/the-di…

Was a pleasure speaking with you @davidmcraney #cognitivesurrender #ai

@davidmcraney @rushkoff Cognitive surrender FTW! @davidmcraney @gidin

Huh, this is now the most downloaded episode since my interview with @rushkoff about billionaire preppers (episode 242). youarenotsosmart.com/2026/04/13/yan…

In @nytopinion “There isn’t a standard definition of A.I. literacy among educators, or a single widely accepted way to measure it,” Jessica Grose writes about her reporting on Seckinger High School in Georgia. nyti.ms/4o8wYxR

#AI can be a valuable tool and, at the same time, it calls for a measured and vigilant approach. The speed and simplicity with which practical assistance can be accessed undoubtedly makes life easier. Yet they can also encourage excessive reliance and the search for ready-made answers, and weaken personal creativity and judgment. #MagnificaHumanitas

Appreciate your thoughts @emollick I am also cool with cognitive surrender in many domains. Can be adaptive and we will lose/transfer some skills that we don’t need. Worth some attention given all the AI doomerism. #cognitivesurrender #ai #deskilling

gm the phrase "cognitive surrender" is making the rounds — people noticing that blindly accepting whatever the model spits out is its own kind of giving up the trick was never to stop thinking. it was to think about different things

"Cognitive surrender is when you stop thinking altogether and blindly accept the answer the AI gives you"

The trend of attorneys getting caught citing AI-hallucinated cases points to a broader problem: instead of checking AI’s work, people keep trusting it spklr.io/6012EMJzk

Canada can be breathtaking. But even the Canadian Rockies have their limits. ⛰️ With AI, photos and videos can look real, even when they are not. #ThinkBeforeYouShare

Parents are worried AI use in the classroom is parasitic, siphoning their children’s intellectual curiosity, capacity for boredom, and data — all in order to make money. nymag.com/intelligencer/…