 |
|
|
|||||
Have you ever had the experience of rereading a sentence multiple times only to realize you still dont understand it? As taught to scores of incoming college freshmen, when you realize youre spinning your wheels, its time to change your approach. This process, becoming aware of something not working and then changing what youre doing, is the essence of metacognition, or thinking about thinking. Its your brain monitoring its own thinking, recognizing a problem, and controlling or adjusting your approach. In fact, metacognition is fundamental to human intelligence and, until recently, has been understudied in artificial intelligence systems. My colleagues Charles Courchaine, Hefei Qiu, Joshua Iacoboni, and I are working to change that. Weve developed a mathematical framework designed to allow generative AI systems, specifically large language models like ChatGPT or Claude, to monitor and regulate their own internal cognitive processes. In some sense, you can think of it as giving generative AI an inner monologue, a way to assess its own confidence, detect confusion, and decide when to think harder about a problem. Why machines need self-awareness Todays generative AI systems are remarkably capable but fundamentally unaware. They generate responses without genuinely knowing how confident or confused their response might be, whether it contains conflicting information, or whether a problem deserves extra attention. This limitation becomes critical when generative AIs inability to recognize its own uncertainty can have serious consequences, particularly in high-stakes applications such as medical diagnosis, financial advice, and autonomous vehicle decision-making. For example, consider a medical generative AI system analyzing symptoms. It might confidently suggest a diagnosis without any mechanism to recognize situations where it might be more appropriate to pause and reflect, like These symptoms contradict each other or This is unusual, I should think more carefully. Developing such a capacity would require metacognition, which involves both the ability to monitor ones own reasoning through self-awareness and to control the response through self-regulation. Inspired by neurobiology, our framework aims to give generative AI a semblance of these capabilities by using what we call a metacognitive state vector, which is essentially a quantified measure of the generative AIs internal cognitive state across five dimensions. 5 dimensions of machine self-awareness One way to think about these five dimensions is to imagine giving a generative AI system five different sensors for its own thinking. Emotional awareness, to help it track emotionally charged content, which might be important for preventing harmful outputs. Correctness evaluation, which measures how confident the large language model is about the validity of its response. Experience matching, where it checks whether the situation resembles something it has previously encountered. Conflict detection, so it can identify contradictory information requiring resolution. Problem importance, to help it assess stakes and urgency to prioritize resources. We quantify each of these concepts within an overall mathematical framework to create the metacognitive state vector and use it to control ensembles of large language models. In essence, the metacognitive state vector converts a large language models qualitative self-assessments into quantitative signals that it can use to control its responses. For example, when a large language models confidence in a response drops below a certain threshold, or the conflicts in the response exceed some acceptable levels, it might shift from fast, intuitive processing to slow, deliberative reasoning. This is analogous to what psychologists call System 1 and System 2 thinking in humans Conducting an orchestra Imagine a large language model ensemble as an orchestra where each musician an individual large language model comes in at certain times based on the cues received from the conductor. The metacognitive state vector acts as the conductors awareness, constantly monitoring whether the orchestra is in harmony, whether someone is out of tune, or whether a particularly difficult passage requires extra attention. When performing a familiar, well-rehearsed piece, like a simple folk melody, the orchestra easily plays in quick, efficient unison with minimal coordination needed. This is the System 1 mode. Each musician knows their part, the harmonies are straightforward, and the ensemble operates almost automatically. But when the orchestra encounters a complex jazz composition with conflicting time signatures, dissonant harmonies, or sections requiring improvisation, the musicians need greater coordination. The conductor directs the musicians to shift roles: Some become section leaders, others provide rhythmic anchoring, and soloists emerge for specific passages. This is the kind of system were hoping to create in a computational context by implementing our framework, orchestrating ensembles of large language models. The metacognitive state vector informs a control system that acts as the conductor, telling it to switch modes to System 2. It can then tell each large language model to assume different rolesfor example, critic or expertand coordinate their complex interactions based on the metacognitive assessment of the situation. Impact and transparency The implications extend far beyond making generative AI slightly smarter. In health care, a metacognitive generative AI system could recognize when symptoms dont match typical patterns and escalate the problem to human experts rather than risking misdiagnosis. In education, it could adapt teaching strategies when it detects student confusion. In content moderation, it could identify nuanced situations requiring human judgment rather than applying rigid rules. Perhaps most importantly, our framework makes generative AI decision-making more transparent.Instead of a black box that simply produces answers, we get systems that can explain their confidence levels, identify their uncertainties, and show why they chose particular reasoning strategies. This interpretability and explainability is crucial for building trust in AI systems, especially in regulated industries or safety-critical applications. The road ahead Our framework does not give machines consciousness or true self-awareness in the human sense. Instead, our hope is to provide a computational architecture for allocating resources and improving responses that also serves as a first step toward more sophisticated approaches for full artificial metacognition. The next phase in our work involves validating the framework with extensive testing, measuring how metacognitive monitoring improves performance across diverse tasks, and extending the framework to start reasoning about reasoning, or metareasoning. Were particularly interested in scenarios where recognizing uncertainty is crucial, such as in medical diagnoses, legal reasoning, and generating scientific hypotheses. Our ultimate vision is generative AI systems that dont just process information but understand their cognitive limitations and strengths. This means systems that know when to be confident and when to be cautious, when to think fast and when to slow down, and when theyre qualified to answer and when they should defer to others. Ricky J. Sethi is a professor of computer science at Fitchburg State University and Worcester Polytechnic Institute. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
For designers of the built environment, it’s necessary to take a long view. Years or even decades can go into the design and construction of a single project, and the best built projects can stand for centuries. But the business of designing buildings is also subject to the upheavals and uncertainty of any given moment, including this very tumultuous one. Looking ahead to the (relatively) short-term future of the next year, Fast Company asked architects from some of the top firms working in the U.S. and around the world to predict the biggest forces shaping the industry this year, and the potential bright spots they might see. Here’s the question we put to a panel of designers and leaders in architecture: What challenges do you see architects tasked with solving in 2026, and what are potential new opportunity areas? Collaboration is key Affordable housing and supporting community resources are in crisisprojects that deliver proximity to public transportation, social infrastructure, and offer cultural resources such as restaurants and entertainment will be in high demand. Understanding the role a building plays within a broader community is a vital part of the design process that is often lacking. Collaboration needs to extend beyond cities and design teams to integrate community needs. This year will bring many of the same challenges we have already seen: more pressure to deliver projects faster while maintaining the quality of the work, understanding what a high-performance building actually means, and streamlining public agency approvals. The latter is an area where AI would be a valuable tool to support innovation and efficiency. There is also a growing opportunity for greater partnership and collaboration with academia and architectural practice. It is important that there is heightened collaboration between the two, particularly because the skill sets of architects are expanding [to include] different job descriptions and needs.Nick Leahy, co-CEO and executive director, Perkins Eastman Resilience is a given 2026 is the year when designing for resilience becomes a given. Innovation will be as much about systems as function, form, and aesthetics. We will think more about embodied carbon, and derive ways to deliver low-carbon buildings without cost premiums. Clients will no longer accept “green is more expensive.” Opportunities: Reuse and reinventionthe second life of a building or district. Conversion of outmoded office buildings to residential and hotels where practical and possible, particularly with older, charming office stock in places where people want to live. Meanwhile, new office buildings will be A++ “luxury,” designed with new forms of amenities centered on wellness and socialization. In the suburbs, malls can become places where mixed-used districts arise, transformed into incubator or civic spaces, designed around health and wellness. Parking lots can be filled with characterful streets and special 24/7 precincts. Workforce housing will also be a big opportunity that fills the gap between luxury and market rate, while data and energy projects will be relevant and exciting for architects not for their novelty but rather for the spatial intelligence and thoughtful planning required in their successful realization.Trent Tesch, principal, KPF Sustainable design is harder than ever One of the most significant challenges facing the U.S. building market in 2026 will be maintaining momentum for sustainable and regenerative design solutions amid economic and policy headwinds. The U.S. construction market has always been driven by a first-cost first mentality, while sustainable design has held its promise of return on investment in the long life cycle of buildings. The hurdle has always been there, but now the bar is even higher with changes to the Energy Star program, the cutting of federal grants for clean energy, reductions to climate resilience programs, and more. So, architects and designers must move beyond purely ROI and well-being conversations to demonstrate how sustainability mitigates risk, ensures compliance, and drives long-term financial resilience.David Polzin, executive director of design, CannonDesign Economic headwinds At PAU we are continuing to incorporate artificial intelligence in aspects of our workflow, but only to augmentnever to replaceour teams talent and judgment. In 2026, architects will probably continue to face economic headwinds. The strong pace of firm consolidation through mergers and acquisitions continues, leaving the question of whether someday it will largely be a discipline split between boutique practices and behemoth corporations.Vishaan Chakrabarti, founder, PAU More than just buildings In 2026 climate volatility, housing inequity, infrastructural breakdown, and economic uncertainty will no longer be background conditions but active forces shaping every decision an architect makes. We will be asked to do more than deliver buildings; we will be expected to repair trust in systemspolitical and economicthat have too often failed communities and the environment. We must navigate these higher expectations, delivering projects with tangible social, environmental, and economic benefits while grappling with tighter timelines and fewer resources. The central challenge will be remaining responsible to both environmental and civic ideals within delivery models that are not designed to reward either.Claire Weisz, founding principal, WXY architecture + urban design Better decisions, earlier Architects are working in a moment where pressure is coming from all sides; climate risks are intensifying, housing affordability remains unresolved, and the industry is still constrained by limited labor and capacity. At the same time, clients increasingly expect early, data-backed answers that show how a design will meet sustainability goals and deliver on long-term building performance outcomes. The challenge is no longer just designing well but navigating increasing complexity and trade-offs without slowing projects down. This is driving the need to remove fragmentation of information across teams and project phases. The defining challenge that architects and designers will need to solve for in 2026 is making confident, defensible decisions early, when they have the biggest impact on [how] a projects environmental, cost, schedule, and performance outcomes are determined.Amy Bunszel, EVP of architecture, engineering, and construction solutions, Autodesk
Category:
E-Commerce
They lie. Repeatedly. Shamelessly. They lie even when the truth would be easier. They lie when the lie can easily be debunked. They lie to dominate, confuse, and assert control. They treat contradiction as an attack and disagreement as betrayal. These are defining traits of narcissistic leadership. Strangely enough, in politics and in organizations alike, we keep rewarding narcissistic leaders by giving them more power. We promote them, fund them, vote for them, excuse them, and normalize their behavior, even when there are unmistakable warning signs that should stop us from doing so. It is obvious that narcissists seek power. The big (and more burning) question is: Why do we keep giving it to them? {"blockType":"mv-promo-block","data":{"imageDesktopUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2026\/01\/PhotoLVitaud-169.jpg","imageMobileUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2026\/01\/PhotoLVitaud-11.jpg","eyebrow":"","headline":"\u003Cstrong\u003ESubscribe to Laetitia@Work\u003C\/strong\u003E","dek":"Women power the worlds productivity its time we talked more about it. Explore a woman-centered take on work, from hidden discrimination to cultural myths about aging and care. Dont miss the next issue subscribe to Laetitia@Work.","subhed":"","description":"","ctaText":"Learn More","ctaUrl":"http:\/\/laetitiaatwork.substack.com","theme":{"bg":"#2b2d30","text":"#ffffff","eyebrow":"#9aa2aa","subhed":"#ffffff","buttonBg":"#3b3f46","buttonHoverBg":"#3b3f46","buttonText":"#ffffff"},"imageDesktopId":91472264,"imageMobileId":91472265,"shareable":false,"slug":""}} We choose narcissists when were anxious Narcissism is often confused with confidence, ambition, or charisma. In reality, pathological narcissism is defined by grandiosity, a constant need for admiration, low empathy, intolerance of criticism, and a tendency to instrumentalize others. At high doses, narcissism is deeply corrosive. Highly narcissistic leaders take greater risks, manipulate more freely, break rules more readily, and do not learn from failure. They externalize blame, rewrite history, and prefer loyal sycophants over competent professionals. As organizational psychologist Adam Grant has argued, we are rarely naive about narcissistic leaders. Most of the time, we recognize them quickly. They boast. They monopolize attention. They perform outrage. They lie openly and repeatedly. We see itand we still choose them. One of the main reasons is that chaos makes us crave certainty. In moments of crisiseconomic instability, war, technological disruption, climate anxietywe mistake loud confidence for competence. Nuance feels weak. Complexity feels unbearable. Fear narrows our tolerance for ambiguity. It makes us vulnerable to leaders who promise control, simplicity, and absolute answersno matter how fictional those answers may be. Seen through this lens, Donald Trump is not really an anomaly. He is a symptom. His constant lying, grandiosity, and contempt for institutions are extreme, but the underlying dynamic is familiar. The same behaviorson a smaller scaleare rewarded every day in companies, startups, media organizations, and public institutions around the world. 7 Things We Must Change If We Want Fewer Narcissistic Leaders If narcissistic leaders keep rising, it is because our systems keep selecting and protecting them. Changing outcomes requires changing the rules of the game. Here are seven shifts that matter. 1. Stop confusing visibility with value Narcissistic leaders thrive on attention. They dominate meetings, interrupt others, and flood the space with what appears to be certainty. In too many environments, visibility is mistaken for contribution. To counter this, organizations must actively redesign how influence is expressedby limiting airtime and prioritizing written input, for example. Value should be measured by clarity created, not noise produced. Treating visibility as value creates a moral hazard: Those least constrained by doubt gain disproportionate influence. 2. Make lying costly Narcissists lie because it works. Lies are tolerated, minimized, or reframed as communication style. This tolerance is fatal. False statements must be corrected publicly and promptly. Repeated dishonesty should carry clear reputational and career consequences. Treating truth as optional corrodes institutions fast. The longer a lie goes unchallenged, the more it signals that reality is negotiableand that power, not truth, sets the terms. 3. Evaluate leaders on collective outcomes Narcissistic leaders often look impressive on individual metrics while quietly hollowing out their teams. Measuring leadership without accounting for turnover, burnout, disengagement, and loss of trust is profoundly wrong. Collective intelligence, psychological safety, and learning capacity must be treated as core performance indicatorsnot soft, secondary concerns. If results are achieved at the expense of trust, retention, and learning, they represent short-term extraction rather than sustainable performance. 4. Stop rewarding the will to power Aggressively wanting power is not proof of leadership potential. In fact, narcissistic personalities are statistically more likely to self-nominate, campaign for authority, and pursue promotion relentlessly. Systems that equate ambition with suitability all but guarantee poor outcomes. Leadership selection should deliberately include capable individuals who do not seek power for its own sakeand should treat excessive self-promotion as a risk signal. 5. Institutionalize dissent Narcissistic leaders fear contradiction and punish it, directly or indirectly. That is why dissent cannot rely on individual bravery alone. Organizations must structurally protect disagreement through formal devils advocate roles, strong whistleblower protections, and explicit rewards for surfacing bad news early. A leader who cannot tolerate dissent is fundamentally dangerous. Disagreement should be seen as a contribution to intelligence. 6. Redefine charisma Charisma is too often equated with dominance, theatrical confidence, and verbal force. But sustainable leadership can look different: calm authority, restraint, curiosity, and the ability to change ones mind in light of new evidence. As long as we glamorize the worst kind of strong personalities, narcissistic leaders will continue to thrive. Our dominant definition of charisma is also deeply gendered. Traits coded as charismaticassertiveness, verbal dominance, emotional detachment, physical presencemap closely onto traditionally masculine norms, while behaviors more often associated with women (like listening) are systematically undervalued. 7. Address the root cause: Fear Narcissistic leaders rise fastest in anxious systems. When people feel unsafeeconomically, socially, psychologicallythey outsource certainty to those who project it most loudly. Reducing precarity, increasing fairness, and building real psychological safety are not just mral imperatives. They are structural defenses against narcissistic leadership. Narcissistic leaders do not seize power alone. They are enabledby our fears, our metrics, our myths about leadership, and our reluctance to confront uncomfortable truths. If we want different leaders, we must become different selectors. The problem is not that narcissists exist. Its that we keep mistaking them for leaders. {"blockType":"mv-promo-block","data":{"imageDesktopUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2026\/01\/PhotoLVitaud-169.jpg","imageMobileUrl":"https:\/\/images.fastcompany.com\/image\/upload\/f_webp,q_auto,c_fit\/wp-cms-2\/2026\/01\/PhotoLVitaud-11.jpg","eyebrow":"","headline":"\u003Cstrong\u003ESubscribe to Laetitia@Work\u003C\/strong\u003E","dek":"Women power the worlds productivity its time we talked more about it. Explore a woman-centered take on work, from hidden discrimination to cultural myths about aging and care. Dont miss the next issue subscribe to Laetitia@Work.","subhed":"","description":"","ctaText":"Learn More","ctaUrl":"http:\/\/laetitiaatwork.substack.com","theme":{"bg":"#2b2d30","text":"#ffffff","eyebrow":"#9aa2aa","subhed":"#ffffff","buttonBg":"#3b3f46","buttonHoverBg":"#3b3f46","buttonText":"#ffffff"},"imageDesktopId":91472264,"imageMobileId":91472265,"shareable":false,"slug":""}}
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||