AI：独立宣言。信任是否是AI无限潜能的边界？

AI: A Declarationof Autonomy Is trust the limit ofAI’s limitless possibilities? AI: A Declarationof Autonomy Welcome to our Technology Vision for 2025. This25th edition of our annual technology trends reportarrives at a watershed moment for technology andhumanity. As more and more leaders embrace theneed to continuously reinvent using tech, dataand AI, they, now more than ever, need a deepunderstanding of AI. Why? Because the rate of AI’s enabler to acting autonomously on behalf ofpeople--equipping them with the capability toperform new tasks and perform others better thanever. Consider the possibilities and opportunities toreinvent as AI finds its way into new and unfamiliarterritories. To truly understand and take advantageof this potential, enterprises will be creating their We believe we can. We see this new age oftechnology as an opportunity to inject trust inAI in a systematic manner so that businessesand people can realize its incredible reinvention Is trust the limit of AI'slimitless possibilities? Julie SweetChair and CEO Karthik NarainGroup Chief Executive We view AI as the new digital because, like digital,it is both a technology and a new way of working.We believe it will be used in every part of theenterprise and it will have a network effect oneverything and everyone involved. Its impact Leaders are aware of the challenges to creatingthis future, which include high up-frontinvestments in their core technologies, data Our research finds that 77% of executives believeunlocking the true benefits of AI will only bepossible when it’s built on a foundation of trust.Leaders must build trust in digital systems andthe AI models, with customers and the workforceby ensuring accuracy, predictability, consistency Accenture research shows that only 36% ofexecutives say their organizations have scaledgen AI solutions, and just 13% report achievingsignificant enterprise-level impact. We are This year’s Technology Vision explores a futurewhen AI transitions from being an automation Contents IntroductionAI: A Declarationof Autonomy 03When LLMs gettheir Bodies 01 The Binary Big Bang 02 Your Face, in the Future 04 The New Learning Loop Differentiating when every interfacelooks the same How people and AI are defininga virtuous cycle of learning, leading, Is trust the limit of AI'slimitless possibilities? How foundation modelsreinvent robotics When AI expands exponentially,systems are upended Page 47-58 Page 22-33 Page 34-46 AI: A Declarationof Autonomy on Artificial General Intelligence (AGI).2,3And likebefore, the race has captivated business leaders, Is trust the limit of AI'slimitless possibilities? But it’s a red herring—a distraction most businessleaders can’t afford. Someday, AGI will be hugelyconsequential, but today it’s still far away with deeptechnical and ethical challenges to address. Instead, We are entering a new chapterin technology—one shaped bya generalization of AI. Today’sproliferation of accessible andever-present AI will drive newlevels of autonomy all throughoutthe business, evolving the ability The Generalization of AI To understand this generalization of AI, one simplyneeds to look around and see how rooted AI isbecoming in our lives. It’s been nearly 30 yearssince Kasparov’s game, and now models that couldmake Deep Blue look like an average player aresitting in everyone’s pockets. The Turing Test, onceconsidered the loftiest benchmark for machineintelligence, is effectively broken every day byconversations people have with Large LanguageModel (LLM)-backed customer service bots andsales agents. Today’s AI models have shrugged off The rush of an AI race is undeniable. We’ve seen it before. In 1997, Garry Kasparovlost a six-game chess match against IBM’s DeepBlue.1It was the first time a computer ever beata chess grandmaster, after decades of testingmachine capability against humans using this Technology Vision 2025| AI: A Declaration of Autonomy society will uplift the world to the next level ofcapability, performance, and progress. It will spuran evolution towards a world enhanced at all levelsby AI cognition, and generate an unprecedented What will the world look like as wide proliferationbrings it to every dimension of our lives? It seemsinevitable that as leaders start to combine theirAI generalization efforts, they will soon enhance industries,it may look like the common frameworkand communications protocol between companiesin an industry, or engines codifying the grand is diffusing across every dimension of our lives,instantly accessible, and—effectively—always there. This is the real disruption to focus on. Because rightnow, even as executives race to implement this newgeneration of AI, few are looking past the separatepieces to truly understand the scope of what they challenges that shape an industry—models thatwill help grow our understanding of things likephysics, genetics, movement, and more. And forcountriesandgovernments, it brings toge