 |
|
|
|||||
Large language models feel intelligent because they speak fluently, confidently, and at scale. But fluency is not understanding, and confidence is not perception. To grasp the real limitation of todays AI systems, it helps to revisit an idea that is more than two thousand years old. In The Republic, Plato describes the allegory of the cave: prisoners chained inside a cave can only see shadows projected on a wall. Having never seen the real objects casting those shadows, they mistake appearances for reality, and they are deprived from experiencing the real world. Large language models live in a very similar cave. LLMs dont perceive the world: they read about it LLMs do not see, hear, touch, or interact with reality. They are trained almost entirely on text: books, articles, posts, comments, transcripts, and fragments of human expression collected from across history and the internet. That text is their only input. Their only experience. LLMs only see shadows: texts produced by humans describing the world. Those texts are their entire universe. Everything an LLM knows about reality comes filtered through language, written by people with varying degrees of intelligence, honesty, bias, knowledge, and intent. Text is not reality: it is a human representation of reality. It is mediated, incomplete, biased, and wildly heterogeneous, often distorted. Human language reflects opinions, misunderstandings, cultural blind spots, and outright falsehoods. Books and the internet contain extraordinary insights, but also conspiracy theories, propaganda, pornography, abuse, and sheer nonsense. When we train LLMs on all the text, we are not giving them access to the world. We are giving them access to humanitys shadows on the wall. This is not a minor limitation. It is the core architectural flaw of current AI. Why scale doesnt solve the problem The prevailing assumption in AI strategy has been that scale fixes everything: more data, bigger models, more parameters, more compute. But more shadows on the wall do not equal reality. Because LLMs are trained to predict the most statistically likely next word, they excel at producing plausible language, but not at understanding causality, physical constraints, or real-world consequences. This is why hallucinations are not a bug to be patched away, but a structural limitation. As Yann LeCun has repeatedly argued, language alone is not a sufficient foundation for intelligence. The shift toward world models This is why attention is increasingly turning toward world models: systems that build internal representations of how environments work, learn from interaction, and simulate outcomes before acting. Unlike LLMs, world models are not limited to text. They can incorporate time-series data, sensor inputs, feedback loops, ERP data, spreadsheets, simulations, and the consequences of actions. Instead of asking What is the most likely next word?, they ask a far more powerful question: What will happen if we do this? What this looks like in practice For executives, this is not an abstract research debate. World models are already emerging (often without being labeled as such), in domains where language alone is insufficient. Supply chains and logistics: A language model can summarize disruptions or generate reports. A world model can simulate how a port closure, fuel price increase, or supplier failure propagates through a network, and test alternative responses before committing capital. Insurance and risk management: LLMs can explain policies or answer customer questions. World models can learn how risk actually evolves over time, simulate extreme events, and estimate cascading losses under different scenarios, something no text-only system can reliably do. Manufacturing and operations: Digital twins of factories are early world models. They dont just describe processes; they simulate how machines, materials, and timing interact, allowing companies to predict failures, optimize throughput, and test changes virtually before touching the real system. /ul> In all these cases, language is useful, but insufficient. Understanding requires a model of how the world behaves, not just how people talk about it. The post-LLM architecture This does not mean abandoning language models. It means putting them in their proper place. In the next phase of AI: LLMs become interfaces, copilots, and translators World models provide grounding, prediction, and planning Language sits on top of systems that learn from reality itself In Platos allegory, the prisoners are not freed by studying the shadows more carefully: they are freed by turning around and confronting the source of those shadows, and eventually the world outside the cave. AI is approaching a similar moment. The organizations that recognize this early will stop mistaking fluent language for understanding and start investing in architectures that model their own reality. Those companies wont just build AI that talks convincingly about the world: theyll build AI that actually understands how it works. Will your company understand this? Will your company be able to build its world model?
Category:
E-Commerce
Valentino, who died on Monday at 93, leaves a lasting legacy full of celebrities, glamour and, in his words, knowing what women want: to be beautiful. The Italian fashion powerhouse has secured his dream of making a lasting impact, outliving Karl Lagerfeld and Yves Saint Laurent. Valentino was known for his unique blend between the bold and colourful Italian fashion and the elegant French haute couturethe highest level of craftsmanship in fashion, with exceptional detail and strict professional dressmaking standards. The blending of these styles to create the signature Valentino silhouette made his style distinctive. Valentinos style was reserved, and over his career, he built upon the haute couture skills he had developed, maintaining his signature style while he led his fashion house for five decades. But he was certainly not without his own controversial views on beauty for women. Becoming the designer Born in Voghera, Italy, in 1932, Valentino Clemente Ludovico began his career early, knowing from a young age he would pursue fashion. He drew from a young age and studied fashion drawing at Santa Marta Institute of Fashion Drawing in Milan before honing his technical design skills at École de la Chambre Syndicale de la Couture Parisienne, the fashion trade association, in Paris. He started his fashion career at two prominent Parisian haute couture houses, first at Jean Desss before moving to Guy Laroche. He opened his own fashion house in Italy in 1959. His early work had a heavy French influence with simple, clean designs and complex silhouettes and construction. His early work had blocked colour and more of a minimalist approach, before his Italian culture really came through later in his collections. He achieved early success through his connections to the Italian film industry, including dressing Elizabeth Taylor fresh off her appearance in Cleopatra (1963). Elizabeth Taylor wearing Valentino while dancing with Kirk Douglas at the party in Rome for the film Spartacus. [Photo: Keystone/Getty Images] Valentino joined the world stage on his first showing at the Pritti Palace in Florence in 1962. His most notable collection during that era was in 1968 with The White Collection, a series of A-line dresses and classic suit jackets. The collection was striking: all in white, while Italy was all about colour. He quickly grew in international popularity. He was beloved by European celebrities, and an elite group of women who were willing to spend the moneythe dresses ran into the thousands of dollars. In 1963, he travelled to the United States to attract Hollywood stars. The Valentino woman Valentinos wish was to make women beautiful. He certainly attracted the A-list celebrities to do so. The Valentino woman was one who would hold themselves with confidence and a lady-like elegance. Valentino wanted to see women attract attention with his classic silhouettes and balanced proportions. Valentino dressed women such as Jackie Kennedy, Audrey Hepburn, Julia Roberts, Gwyneth Paltrow, and Anne Hathaway. Anne Hathaway and Valentino Garavani attend the 2011 Oscars. [Photo: Fairchild Archive/Penske Media/Getty Images] His aristocratic taste inherited ideas of beauty and old European style, rather than innovating with new trends. His signature style was formal designs that had the ability to quietly intimidateincluding the insatiable Valentino red. Red was a signature colour of his collections. The colour provided confidence and romance, while not distracting away from the beauty of the woman. French influence Being French-trained, Valentino was well acquainted with the rules of couture. With this expertise, he was one of the first Italian designers to be successful in France as an outsider with the launch of his first Paris collection in 1975. This Paris collection showcased more relaxed silhouettes with many layers, playing towards the casual nature of fashion. A model in the Valentino Spring 1976 ready to wear collection walks the runway in Paris in 1975. [Photo: Fairchild Archive/WWD/Penske Media/Getty Images] While his design base was in Rome, many of his collections were shown in Paris over the next four decades. His Italian culture mixed with the technicality of Parisian haute couture made Valentino the designer he was. Throughout his career, his designs often maintained a classic silhouette bust, matched with a bold Italian colour or texture. Unlike some designers today, Valentinos collections didnt change too dramatically each season. Instead, they continued to maintain the craftsmanship and high couture standards. Quintessentially beautiful is often the description of Valentinos work however this devotion to high beauty standards has seen criticism of the industry. In 2007, Valentino defended the trend of very skinny women on runways, saying when girls are skinny, the dresses are more attractive. Critics said his designs reinforce exclusion, gatekeeping fashion from those who dont conform to traditional beauty standards. The Valentino runways only recently have started to feature more average sized bodies and expand their definition of beauty. The $300 million sale of Valentino The Valentino fashion brand sold for US$300 million in 1998 to Holding di Partecipazioni Industriali, with Valentino still designing until his retirement in 2007. Valentino sold to increase the size of his brand: he knew without the support of a larger corporation surviving alone would be impossible. Since Valentinos retirement, the fashion house has continued under other creative directors. Valentino will leave a lasting legacy as the Italian designer who managed to break through the noise of the French haute couture elite and make a name for himself. The iconic Valentino red will forever be remembered for its glamour, and will live on with his legacy. A true Roman visionary with unmatched craftsmanship. Jye Marshall is a lecturer of fashion design at the School of Design and Architecture at Swinburne University of Technology. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
In 2019, during his first term, U.S. President Donald Trump expressed a desire to buy Greenland, which has been a part of Denmark for some 300 years. Danes and Greenlanders quickly rebuffed the offer at the time. During Trumps second term, those offers have turned to threats. Trump said on his social media platform Truth Social in late December 2024 that, for purposes of national security, U.S. control over Greenland was a necessity. The president has continued to insist on the national security rationale into January 2026. And he has refused to rule out the use of military force to control Greenland. From my perspective as an international relations scholar focused on Europe, Trumps national security rationale doesnt make sense. Greenland, like the U.S., is a member of NATO, which provides a collective defense pact, meaning member nations will respond to an attack on any alliance member. And because of a 1951 defense agreement between the U.S. and Denmark, the U.S. can already build military installations in Greenland to protect the region. Trumps 2025 National Security Strategy, which stresses control of the Western Hemisphere and keeping China out of the region, provides insight into Trumps thinking. US interests in Greenland The United States has tried to acquire Greenland several times. In 1867, Secretary of State William Seward commissioned a survey of Greenland. Impressed with the abundance of natural resources on the island, he pushed to acquire Greenland and Iceland for US$5.5 millionroughly $125 million today. But Congress was still concerned about the purchase of Alaska that year, which Seward had engineered. It had seen Alaska as too cold and too distant from the rest of the U.S. to justify spending $7.2 millionroughly $164 million todayalthough Congress ultimately agreed to do it. There was not enough national support for another frozen land. In 1910, the U.S. ambassador to Denmark proposed a complex trade involving Germany, Denmark and the United States. Denmark would give the U.S. Greenland, and the U.S. would give Denmark islands in the Philippines. Denmark would then give those islands to Germany, and Germany would return Schleswig-HolsteinGermanys northernmost stateto Denmark. But the U.S. quickly dismissed the proposed trade as too audacious. During World War II, Nazi Germany occupied Denmark, and the U.S. assumed the role of protector of both Greenland and Iceland, both of which belonged to Denmark at the time. The U.S. built airstrips, weather stations and radar and communications stationsfive on Greenlands east coast and nine on the west coast. The Pituffik Space Base, formerly Thule Air Base, is pictured in northern Greenland on Oct. 4, 2023. [Photo: Thomas Traasdahl/Ritzau Scanpix/AFP via Getty Images] The U.S. used Greenland and Iceland as bases for bombers that attacked Germany and German-occupied areas. Greenland had a high value for military strategists because of its location in the North Atlanticto counter Nazi threats to Allied shipping lanes and protect transatlantic routes, and because it was a midpoint for refueling U.S. aircraft. Greenlands importance also rested on its deposits of cryolite, useful for making aluminum. In 1946, the Truman administration offered to buy Greenland for $100 million, as U.S. military leaders thought it would play a critical role in the Cold War. The secret U.S. project Operation Blue Jay at the beginning of the Cold War resulted in the construction of Thule Air Base in northwestern Greenland, which allowed U.S. bombers to be closer to the Soviet Union. Renamed Pituffik Space Base, today it provides a 24/7 missile warning and space surveillance facility that is critical to NATO and U.S. security strategy. At the end of World War II, Denmark recognized Greenland as one of its territories. In 1953, Greenland gained constitutional rights and became a country within the Kingdom of Denmark. Greenland was assigned self-rule in 1979, and by 2009 it became a self-governing country, still within the Kingdom of Denmark, which includes Denmark, Greenland and the Faroe Islands. Denmark recognizes the government of Greenland as an equal partner and recently gave it a more significant role as the first voice for Denmark in the Arctic Council, which promotes cooperation in the Arctic. What the US may want The Trump administrations 2025 National Security Strategy identifies three threats in the Western Hemisphere: migration, drugs and crimes, and Chinas increasing influence. Two of those threats are irrelevant when considering Greenland. Greenlandic people are not migrating to the U.S., and they are not drug traffickers. However, Greenland is rich in rare earth minerals, including neodymium, dysprosium, graphite, copper and lithium. Additionally, China seeks to establish mining interests in Greenland and the Arctic as part of its Polar Silk Road initiative. China had offered to build an infrastructure for Greenland, including improving the airport, until Denmark stepped in and offered airport funding. And China has worked with Australian companies to secure mining opportunities on the island. A U.S. Air Force helicopter flies near Thule Air Base in Greenland in 1955. [Photo: James McAnally/Archive Photos/Getty Images] Those rare earth minerals appeal to the European Union, too. The EU lists some 30 raw materials that are essential for their economies. Twenty-five are in Greenland. The Trump administration has made it clear that controlling these minerals is a national security issue, and the president wants to keep them away from China. Figures vary, but it is estimated that over 60% of rare earth elements or minerals are currently mined in China. China also refines some 90% of rare earths. This gives China tremendous leverage in trade talks. And it results in a dangerous vulnerability for the U.S. and other nation states seeking to modernize their economies. With few suppliers of these rare earth elements, the political and economic costs of securing them are high. Greenland has only two operating mines. One is the Tan Breez project in southern Greenland. It produces 17 metals, including terbium and neodymium, that are used in high-strength magnets used in many green technologies and in aircraft manufacturing, including for the F-35 fighter planes. Consider for a moment that Trump is not interested in owning Greenland. Instead, he is using this threatening position to secure promises from the Greenlandic government to make economic deals with the U.S. and not China. Thus, Trumps threats could be less about national security and much more about eliminating competition from China and securing wealth for U.S. interests. This form of coercive diplomacy threatens the political and economic development of not only Greenland but Europe. In recent interviews, Trump has made it clear that he does not respect international law and the sovereignty of countries. His position, I believe, undermines the international order and removes the U.S. as a responsible leader of that framework established after World War II. Steven Lamy is a professor emeritus of political science and international relations and spatial sciences at USC Dornsife College of Letters, Arts and Sciences. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
Autodesk, Inc., a maker of digital design software, announced Thursday it plans to lay off about 1,000 employees, largely in sales roles. The announcement comes just a week after another tech company, Meta, announced it would eliminate up to 1,500 positions. Heres what you need to know about the latest tech company layoffs. Whats happened? The plan will reduce Autodesk’s workforce by approximately 7%. Autodesk indicated a significant number of the affected jobs would be in customer-facing sales roles. The plan will also reallocate resources to accelerate strategic priorities, The Wall Street Journal reported. CEO Andrew Anagnost assured employees in a letter that these layoffs were not indicative of a yearly pattern or a shift to replacing workers with AI. Last February, Autodesk announced a plan to reduce its workforce by 9%. How many jobs are being lost? The 2026 layoffs are projected to affect around 1,000 employees. Autodesks 2025 layoffs were projected to affect around 1,350 employees, according to Anagnosts internal message. Autodesk predicts pre-tax restructuring charges between $135 and $160 million, largely tied to employee termination benefits. The restructuring plan is expected to conclude by the end of fiscal year 2027. Why is Autodesk laying off employees? This move is part of the final phase in Autodesks recent efforts to optimize sales and marketing. Last Februarys layoffs were informed by the need to drive more efficiency and focus to implement specific programs, according to Anagnosts statement to employees. In the letter to employees, Anagnost said the reduction focuses on completing the companys go-to-market transformation, expanding its AI and platform capabilities and strengthening corporate functions. At Autodesk University in September 2025, the company revealed brand-new AI tools with pre-beta software, indicating a big step forward in Autodesks AI investments. One such tool was a neural CAD, which Autodesk said could automate 80-90% of routine design tasks. Other 2026 tech layoffs Just last week, Meta announced a 10% reduction in its Reality Labs division, the section of the company primarily responsible for augmented and virtual reality (like the metaverse). It was the largest layoff announcement in the tech sector so far in 2026. Nearly 124,000 employees across 269 tech companies were laid off in 2025, according to data compiled by Layoffs.fyi. With Meta and Autodesk’s projected reductions alone, at least 2,500 employees across the two companies will lose their jobs. Since 2023, Autodesk has laid off at least 1,600 employees.
Category:
E-Commerce
After spending much of his career in marketing, Daniel Hebert pivoted into Software as a Service (SaaS) sales for a high-growth startup in 2018. But what started out as a dream job soon turned into a nightmare. Like many tech startups, the business went from growth mode to scaling down as the market turned in 2022. As the head of the sales team, Hebert found himself on the front lines of that cost-cutting campaign. I would be assigned a team and have to fire half the people. Then Id have to rebuild the process and rebuild the team. I did three or four cycles of that in 15 months, and I just got so insanely burnt out, he says. I just randomly started getting [stress-induced] vertigo, to the point where I was bedridden for days at a time. Like, I couldnt stand up without falling over. Though he was just a couple of months shy of a big payday when the company was set to be acquired, Hebert resigned and began offering independent sales coaching services in early 2023. In the very early days, I used AI a lot for brainstorming: figuring out, like, how do I market this thing? he says. Then it became kind of like a marketing editor, and then I started using it with my clients. And from there, it went wild. As he got more comfortable with the technology, Hebert says he started to look for ways he could leverage it to scale his offering. If I want to increase my income as a coach, I have to add more coaching hours, and theres so many hours I can coach before I max out, he says. So if I want to scale my income, I have to do it in a way that doesnt require my time. Hebert says he had lots of ideas for tech products over the years, but never had the time, resources, or capacity to build them. That is, until he became a solopreneur and learned how to use Lovable, an AI-powered app builder. I could actually build functioning B2B SaaS software, and it completely changed everything for me, he says. I could now build a handful of these tools and get some subscribers, and have this portfolio of apps generating income without me having to add more coaching hours. AI is widely expected to be a game changer for businesses of all shapes and sizes, enabling new kinds of productivity and revenue. But that transformation is already well underway among solopreneurs. Thats because those who are independently employed dont have the practical constraints that hold back those operating in traditional organizations, like legal compliance, or a lengthy rollout process of new tools involving large groups of people and many teams. They also enjoy more direct financial rewards from being early AI adopters. Why solopreneurs are winning the AI race In a recent survey conducted by online freelance marketplace Upwork, 90% of freelancers said AI is helping them learn new skills faster, and 88% said it positively impacted their careers. Furthermore, 34% strongly agree that AI gives them an edge, compared with 28% of those who are traditionally employed. Overall, 54% of independently employed professionals self-report advanced AI proficiency, compared with 38% of those who are independently employed. The data overwhelmingly shows that independent talent is adopting AI at higher rates, says Gabby Burlacu, a senior researcher at the Upwork Research Institute and one of the studys authors. They are faster at adopting it, and they are also more proficient in finding ways to use these tools to improve their work, and actually embedding them in their workflows when we compare them to full-time employees. As traditional organizations sort through how to deploy which AI tools and embark on lengthy integration and change management campaigns, solopreneurs are taking it upon themselves to experiment on their own. They are pursuing self-developmentwhich we’re seeing across the boardat much higher rates than full-time employees who tend to rely more on organizational learning and development, Burlacu says. The tools are available to everyone, but the ability to unlock their power and do better work comes from being in control of how you work, how you learn. Thats the game changer. Not just for knowledge workers While knowledge industry workers are the most likely to latch onto the technology, solopreneurs of all shapes and sizes are finding ways to grow their business using AI. According to a study conducted by business insurance provider Simply Business, about 37.5% of independently employed knowledge workers, such as accountants, IT specialists and legal advisers, are using AI in their day-to-day operations. At the same time, so are 29.7% of those with independent creative or lifestyle ventureslike artists, coaches, and photographersand 14% of tradespeople and contractors. Those figures, however, only capture intentional adoption, and don’t include those that are using other software tools to run their business, which are increasingly integrating AI functionality. According to the survey, more than half use digital marketing tools, and another 39% leverage accounting software. Some of those, quite frankly, have AI built in that they may not even realize they’re using, says Dana Edwards, group chief technology officer at Simply Business. In our 2025 survey, 23% [of solopreneurs] were using AI, and if you kind of follow the trend line, I wouldn’t be surprised if we were in the 60 or 70% range in 2026. AI offers solopreneurs direct value In a recent survey of solopreneurs who are already using AI conducted by Zoom and Upwork, 93% agreed or strongly agreed that the technology is critical to their success. Furthermore, 75% said it has reduced costs, 89% said its helped them expand into new markets, 78% said it allowed them to automate repetitive tasks, and 74% said it’s directly contributed to a new product or service offering. Seven out of 10 of those individuals are already seeing revenue impact from incorporating AI in their day-to-day operations, and were just at the start, says Lisa Scheiring, Zooms new global small business advisor and chief solo officer. Technology now is enabling those individuals to build durable, scalable businesses that can compete directly with some of the larger enterprises in their categories. Scheiring says the establishment of her position, along with Zooms recently announced Solopreneur 50 recognition program, are a testament to the impact the company believes businesses of one will have in the AI era. She adds that its not just the lack of traditional constraints that allows solopreneurs to leverage AI more readily; its also about personality and risk appetite. If you are bold enough to take on a solo venture, you are already someone who is willing to learn new things and change and grow. Those are the same skills that you need to learn how to use AI, Scheiring says. Someone who s willing to take that step forward and be a solopreneur, those same capabilitieswhen applied to a new technology like AIwill help them move further, faster. That proved the case for Hebert, who went from burned-out sales professional to independent sales coach to launching his own tech products in less than three years. You used to have to raise a lot of money and hire a big team to do things like thisnow you just need an idea and the patience to actually learn the AI skills to make it happen, he says. Thats the superpower of the solopreneur.
Category:
E-Commerce
Sites : [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] next »