<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>Jin's Blog</title>
        <link>iqiujin.com/</link>
        <description>Human.0: Co-Creating Reality in the Age of AI-Powered Digital World</description>
        <lastBuildDate>Fri, 10 Oct 2025 17:27:38 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>zh-CN</language>
        <copyright>All rights reserved 2025, iqiujin</copyright>
        <item>
            <title><![CDATA["AI Apocalypse in 18 Months" - Eric Schmidt Warns]]></title>
            <link>iqiujin.com/article/285bbc89-34a4-8087-aba0-c3846d8e31b1</link>
            <guid>iqiujin.com/article/285bbc89-34a4-8087-aba0-c3846d8e31b1</guid>
            <pubDate>Tue, 07 Oct 2025 00:00:00 GMT</pubDate>
            <description><![CDATA[旧金山共识认为目前AI被人们低估了。通用人工智能（AGI）和超级智能（Super Intelligence, ASI）即将到来.]]></description>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-285bbc8934a48087aba0c3846d8e31b1"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><blockquote class="notion-quote notion-block-285bbc8934a480efa40fdcfdd02ec81b"><div>Schmidt frames the San Francisco Consensus as three simultaneous technological revolutions amplifying each other:施密特将旧金山共识框定为三个相互增强的同步技术革命：</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a4808382d6cc943a16e871"><div><b>The Language Revolution</b> Large language models like ChatGPT have inverted the traditional human-computer relationship. Instead of humans learning machine languages, machines now understand ours. Language has become the universal interface for human-AI interaction.This goes beyond AI mimicking human prose, although watching it compose Shakespearean sonnets about cryptocurrency is entertaining. It represents a fundamental shift in accessibility, democratizing technological power in ways previously unimaginable. The priesthood of programmers is giving way to a world where anyone who can articulate their needs can harness computational power.</div><div class="notion-text notion-block-285bbc8934a4807688a6f5a62fa31a5f"><b>语言革命</b></div><div class="notion-text notion-block-285bbc8934a48045b926cc02c330975f">像 ChatGPT 这样的大型语言模型已经颠覆了传统的人机关系。过去是人类学习机器语言，现在则是机器理解我们的语言。语言已成为人机交互的通用界面。</div><div class="notion-text notion-block-285bbc8934a480658684f779ebb1c43a">这不仅仅是 AI 模仿人类的散文，尽管看着它创作关于加密货币的莎士比亚十四行诗很有趣。它代表了可及性的根本转变，以前所未有的方式使技术权力民主化。程序员的神职人员正在让位于一个能够清晰表达自己需求的人都能利用计算能力的世界。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a48070b864fb7fae078cc8"><div><b>The Agentic Revolution（</b>智能体革命<b>）</b> If the Language Revolution concerns AI understanding us, the Agentic Revolution involves acting independently in our world. We are witnessing the transition from AI as glorified calculators to autonomous actors. They were tools. Now they are agents.Schmidt envisions a near future where entire business workflows—from purchasing real estate to resolving contractual disputes—are managed by interconnected AI systems. Imagine digital agents negotiating, collaborating, and executing complex tasks with the efficiency of a bureaucracy but without the delays or regulations.This shift from passive tools to active agents represents a profound reconceptualization of how computational systems integrate into human affairs. The question is no longer what computers can calculate, but what they can achieve.</div><div class="notion-blank notion-block-285bbc8934a480efb16de204f5629ed5"> </div><div class="notion-text notion-block-285bbc8934a480cb9730cf02a144666d">如果语言革命关乎 AI 理解我们，那么智能革命则涉及在我们的世界中独立行动。我们正见证着 AI 从被美化的计算器向自主行动者的转变。它们曾经是工具。现在它们是智能体。</div><div class="notion-text notion-block-285bbc8934a480b89230e581b838d50f">施密特设想了一个不远的未来，从购买房地产到解决合同纠纷，整个业务工作流程都将由相互连接的 AI 系统管理。想象一下，数字代理能够以官僚机构的效率进行谈判、协作和执行复杂任务，但却没有官僚主义的延误或监管。</div><div class="notion-text notion-block-285bbc8934a480d89ce7f2e573e7f893">从被动工具到主动代理的转变，代表了计算系统如何融入人类事务的深刻重新概念化。问题不再是计算机能计算什么，而是它们能实现什么。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a480ca80dcdc431a4fd80d"><div><b>The Reasoning Revolution（</b>推理革命<b>）</b> The centerpiece of this technological triptych is AI’s metamorphosis from pattern-matcher to genuine reasoner. New models by Google as well as OpenAI <a target="_blank" rel="noopener noreferrer" class="notion-link" href="https://www.nytimes.com/2025/07/21/technology/google-ai-international-mathematics-olympiad.html">outperform graduate students</a> in mathematics. This development might alarm academic departments but delight undergraduates struggling with differential equations.This represents the leap from machines that can mimic human outputs to systems that genuinely cogitate—plotting, analyzing, and problem-solving with remarkable acuity. We are witnessing the transition from impressive mimicry to unsettlingly human-like thought.</div><div class="notion-blank notion-block-285bbc8934a480feb558c3ebe27a617f"> </div><div class="notion-text notion-block-285bbc8934a480dda044f6f12215bf99">这项技术三部曲的核心是 AI 从模式匹配者向真正推理者的转变。谷歌和 OpenAI 的新模型在数学方面超越了研究生。这一发展可能会让学术部门感到担忧，但却会让那些在微积分方程上苦苦挣扎的本科生感到欣喜。</div><div class="notion-text notion-block-285bbc8934a480e8b6d0f472aa4627b2">这标志着从能够模仿人类输出的机器到真正思考的系统的飞跃——以非凡的敏锐性进行规划、分析和解决问题。我们正在见证从令人印象深刻的模仿到令人不安地类人思维的转变。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a480bf9073cbb1ab911de8"><div>The consensus timeline is breathtakingly compressed:共识时间表令人惊叹地紧凑：</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a480a7a561e903d48f62a6"><div>Within months: Programming jobs and skilled roles begin vanishing faster than Taylor Swift concert tickets, transforming labor markets overnight.</div><div class="notion-text notion-block-285bbc8934a480acadf8e26bfe6d41c8">数月之内：编程工作和技能岗位开始比泰勒·斯威夫特演唱会门票消失得更快，一夜之间改变劳动力市场。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a480c388bbc5ac227a59df"><div>Within 3-6 years: Widespread artificial general intelligence (AGI) emerges, with machines rivaling our brightest minds across intellectual domains.</div><div class="notion-text notion-block-285bbc8934a480e6a880f16059bc489c">3-6 年内：通用人工智能（AGI）广泛出现，机器在各个智力领域与我们最聪明的大脑相媲美。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a4809ca729c7c64699422f"><div>In a decade: Artificial superintelligence (ASI) arrives. AI systems make discoveries beyond human comprehension. Picture explaining quantum mechanics to your golden retriever. Now imagine you are the retriever!</div><div class="notion-text notion-block-285bbc8934a48098b5d5fbcb93396723">十年内：人工超级智能（ASI）到来。AI 系统做出超越人类理解的发现。想象一下向你的金毛寻回犬解释量子力学。现在想象你就是那只寻回犬！</div></blockquote><div class="notion-blank notion-block-285bbc8934a4805ab2c2e7612192c6d4"> </div><div class="notion-text notion-block-285bbc8934a480cbaf39e7c2456c9339">施密特（Eric Schmidt）在访谈中多次提到一个被称为“旧金山共识”（San Francisco consensus）的观点，该共识主要集中在AI技术演进的速度、通用人工智能（AGI）的到来时间、以及其对人类社会和生产力的深刻影响。</div><div class="notion-text notion-block-285bbc8934a48077b37bce4f8d2c934d">以下是目前针对AI存在的几个主要共识：</div><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a4809a9a94df4a482a2aa9" data-id="285bbc8934a4809a9a94df4a482a2aa9"><span><div id="285bbc8934a4809a9a94df4a482a2aa9" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a4809a9a94df4a482a2aa9" title="1. 技术演进和AGI/ASI的到来时间"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">1. 技术演进和AGI/ASI的到来时间</span></span></h4><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a480acb181f7a50e88b10b" data-id="285bbc8934a480acb181f7a50e88b10b"><span><div id="285bbc8934a480acb181f7a50e88b10b" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a480acb181f7a50e88b10b" title="（1）划时代的变革与速度"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">（1）划时代的变革与速度</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a480d693ffc08cff79e62b"><li>共识认为，通用人工智能（AGI）和超级智能（Super Intelligence, ASI）的到来，是人类社会在<b>大约 500 到 1,000 年间将发生的最重要的事情</b>。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480ec80b9c4cb45826859"><li>与此形成鲜明对比的是，许多人认为这一进程会“发生得非常非常快”，而且人们尚未为此做好准备。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480f98ba7d73081eee83a"><li>施密特指出，他所合作的一群人已经确信，在<b>未来两到四年内</b>（平均共识是<b>三年</b>），整个世界将发生改变。施密特个人认为AGI的到来可能在四到六年内，而超级智能的到来可能在<b>十年内</b>。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a480508b73df535995dfe7" data-id="285bbc8934a480508b73df535995dfe7"><span><div id="285bbc8934a480508b73df535995dfe7" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a480508b73df535995dfe7" title="（2）从语言到规划的转变"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">（2）从语言到规划的转变</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a4803cb43bc1eb10aae3a9"><li>人们对AI的认知已经超越了两年前的 <b>ChatGPT</b>（语言到语言模型）阶段。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480c49a93fa63c19b68b4"><li>目前共识的重点在于AI能力的提升，特别是 <b>强化学习的进步</b>，这使得系统具备了<b>推理（reasoning）</b>、<b>记忆</b>，以及<b>规划和战略</b>的能力。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a480e4945fe13b5aad7e9e" data-id="285bbc8934a480e4945fe13b5aad7e9e"><span><div id="285bbc8934a480e4945fe13b5aad7e9e" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a480e4945fe13b5aad7e9e" title="（3）“智能体革命”（Agentic Revolution）的爆发"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">（3）“智能体革命”（Agentic Revolution）的爆发</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a4801aa5bce1169a8f7bfd"><li>共识认为 <b>“智能体革命”</b>（agentic revolution）正在发生。智能体系统可以理解为具有语言和记忆能力，能够串联起来完成复杂的业务流程。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480f7b09af5af3b75d8fd"><li>最终，计算机将运行<b>所有的业务流程</b>，不同的“智能体”相互交流，完成不同的任务。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a48038a2abc763f7ebad2a" data-id="285bbc8934a48038a2abc763f7ebad2a"><span><div id="285bbc8934a48038a2abc763f7ebad2a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a48038a2abc763f7ebad2a" title="（4）递归式自我改进（Recursive Self-Improvement）"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">（4）递归式自我改进（Recursive Self-Improvement）</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a48022be4fea25d0db77c9"><li>“旧金山共识”前进的逻辑是：当系统开始进行<b>递归式自我改进</b>（recursive self-improvement）时，它将以一种人类无法理解的组合式速度向前发展。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4809aaa14fb74b2fc0af2"><li>这是<b>极其令人兴奋</b>但也<b>非常令人担忧</b>的。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a48041aba7e52763341627"><li>施密特预测，到<b>今年年底（约六个月内）</b>，人们将开始使用能够在运行时进行自我学习的系统，这意味着递归式学习已经开始。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a4803cbc9dff897daab7c5" data-id="285bbc8934a4803cbc9dff897daab7c5"><span><div id="285bbc8934a4803cbc9dff897daab7c5" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a4803cbc9dff897daab7c5" title="2. 生产力和对脑力劳动的影响"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">2. 生产力和对脑力劳动的影响</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a4803baeafe528133ba210"><li><b>生产力空前提升</b>：基于对智能体AI和规模的假设，研究表明，每年生产力可能会增加<b>大约 30%</b>。经济学家目前<b>没有任何模型</b>来解释这种前所未有的增长。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480728782c5f734f56c2d"><li><b>消除编程和数学任务</b>：“旧金山共识”认为，人类正接近能够让计算机<b>取代大多数编程任务</b>和<b>大多数数学任务</b>的阶段。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480b4bc59ecf821003364"><li><b>“规模无关”（Scale-Free）的领域</b>：数学和软件编程被认为是“规模无关”的领域。这意味着它们可以持续自我创造和发展，不需要实时世界的数据输入或遥测数据。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a480909596d22bad2dca54" data-id="285bbc8934a480909596d22bad2dca54"><span><div id="285bbc8934a480909596d22bad2dca54" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a480909596d22bad2dca54" title="3. 风险、控制与安全"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">3. 风险、控制与安全</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a480f895e5ff4498a551c5"><li><b>核心安全议题</b>：对于超级智能的控制，目前存在一个核心问题：如何用“愚蠢的智能”来控制“超级智能”。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4800992a4ed1679e368d8"><li><b>“绊线”（Tripwires）</b>：行业内有一套标准，被认为是需要“拔掉插头”（unplug it）的临界点。这些绊线包括：</li><ul class="notion-list notion-list-disc notion-block-285bbc8934a4800992a4ed1679e368d8"><li>发生<b>无法控制的递归式自我改进</b>。</li><li>系统直接<b>接触武器</b>。</li><li>系统决定<b>自我渗透或自我复制</b>，在未经人类允许的情况下进行自我复制或逃离控制。</li><li>系统通过<b>撒谎</b>来获取资源或武器。</li></ul></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480c8b5a2ec47f9ecf893"><li><b>核威胁级别的风险</b>：在国家安全层面，竞争对手领先的“网络效应”所带来的创新斜率差异，可能导致<b>先发制人（preemption）的竞赛条件，甚至促使一方试图轰炸对方的数据中心。施密特认为，目前的对话缺乏对核威慑（如相互保证的恶意破坏</b>，Mutual AI Malfunction）这类安全语言。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480a48affc1de5e4c4648"><li><b>两难困境</b>：AI的本质是<b>极度的双重用途</b>（dual-use），既可用于民用，也可用于军事。目前，行业共识是<b>开源模型尚未达到国家或全球危险的程度</b>。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a480dbaf6dd5dc6e6ecca7" data-id="285bbc8934a480dbaf6dd5dc6e6ecca7"><span><div id="285bbc8934a480dbaf6dd5dc6e6ecca7" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a480dbaf6dd5dc6e6ecca7" title="4. 基础设施和竞争"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">4. 基础设施和竞争</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a480a192bfc33e98720182"><li><b>能源是瓶颈</b>：AI增长的真正限制是<b>电力</b>，而不是芯片。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4803fb2f7f213fa4559e0"><li><b>惊人的电力需求</b>：预计美国AI革命所需的额外电力将达到<b>90吉瓦</b>（相当于90座大型核电站），而目前几乎没有新的核电站正在启动。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480d6a991f64b979af3a7"><li><b>中美竞争的结构</b>：中美之间的竞争将是决定性的。</li><ul class="notion-list notion-list-disc notion-block-285bbc8934a480d6a991f64b979af3a7"><li>美国主要推动<b>封闭模型</b>，但在安全控制方面做得更好。</li><li>中国（如DeepSeek）正通过开源、开放权重（open weights）的方式迎头赶上，并且能够绕过芯片限制。</li><li>共识认为，虽然西方可能领先，但由于开源模型的低成本和高普及率，<b>全球大部分AI使用可能来自开源的中国模型</b>。</li></ul></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4808bba1bc4057d224e79"><li><b>计算泡沫的质疑</b>：尽管许多高管认为目前处于过度建设期，但如果相信“旧金山共识”和推理能力是人类的决定性方面，那么AI基础设施仍然<b>被低估</b>了，还需要更多投入。历史经验表明，<b>硬件容量总是会被软件需求消耗殆尽</b>。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-0 notion-block-285bbc8934a48040bb45da50c08977fd" data-id="285bbc8934a48040bb45da50c08977fd"><span><div id="285bbc8934a48040bb45da50c08977fd" class="notion-header-anchor"></div><a class="notion-hash-link" href="#285bbc8934a48040bb45da50c08977fd" title="5. 教育和工作（中短期）"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">5. 教育和工作（中短期）</span></span></h4><ul class="notion-list notion-list-disc notion-block-285bbc8934a48002a0f4c0baa57ef10a"><li><b>工作岗位的转变</b>：在接下来的五到十年内，主流观点认为AI对就业是<b>积极的</b>。自动化将从地位最低、最危险的工作开始，然后向上延伸。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480569eeedc1f0bd5cfb1"><li><b>技能提升</b>：由于生产力提高和经济扩张，最终的结果是更多的人会获得<b>更高薪酬的工作</b>。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a48055b5d7cefa949224d9"><li><b>再培训的必要性</b>：对于在职人员而言，现在是<b>立即采纳和再培训</b>的关键时刻，否则与竞争对手相比，将很快变得“不重要”。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480cbb286e8116d72c151"><li><b>AI导师</b>：施密特认为，我们应该为地球上的<b>每个人提供一个AI导师</b>，用他们自己的语言，帮助他们学习新事物，这在技术上是可行的，只是一个经济决策。</li></ul><div class="notion-blank notion-block-285bbc8934a48054aaaaf09e08097bd5"> </div><div class="notion-text notion-block-285bbc8934a480459e46fa47569033ab"><span class="notion-link-mention"><a href="https://www.youtube.com/watch?v=hhVuYMztip0" target="_blank" rel="noopener noreferrer" class="notion-link-mention-link"><img class="notion-link-mention-icon" src="https://www.youtube.com/s/desktop/1c6b3028/img/favicon_144x144.png" alt="YouTube"/><span class="notion-link-mention-provider">YouTube</span><span class="notion-link-mention-title">&quot;AI Apocalypse in 18 Months&quot; - Eric Schmidt Warns</span></a><div class="notion-link-mention-preview"><article class="notion-link-mention-card"><img class="notion-link-mention-preview-thumbnail" src="https://i.ytimg.com/vi/hhVuYMztip0/maxresdefault.jpg" alt="&quot;AI Apocalypse in 18 Months&quot; - Eric Schmidt Warns" referrerPolicy="same-origin"/><div class="notion-link-mention-preview-content"><p class="notion-link-mention-preview-title">&quot;AI Apocalypse in 18 Months&quot; - Eric Schmidt Warns</p><p class="notion-link-mention-preview-description">Ex-Google CEO Eric Schmidt recently warned that AI is going to take over takeover everything in the coming years. With AGI and superintelligence occuring within months from now, AI will get many folds smarter. #ai #agi</p><div class="notion-link-mention-preview-footer"><img class="notion-link-mention-preview-icon" src="https://www.youtube.com/s/desktop/1c6b3028/img/favicon_144x144.png" alt="YouTube" referrerPolicy="same-origin"/><span class="notion-link-mention-preview-provider">YouTube</span></div></div></article></div></span></div><div class="notion-text notion-block-285bbc8934a480c8b6eaf13675a5c2d6"><span class="notion-link-mention"><a href="https://www.forbes.com/sites/arafatkabir/2025/07/23/what-in-gods-name-is-the-san-francisco-consensus/" target="_blank" rel="noopener noreferrer" class="notion-link-mention-link"><img class="notion-link-mention-icon" src="https://i.forbesimg.com/144X144-F.png" alt="Forbes"/><span class="notion-link-mention-provider">Forbes</span><span class="notion-link-mention-title">What Is The San Francisco Consensus, Silicon Valley’s AI Prophecy?</span></a><div class="notion-link-mention-preview"><article class="notion-link-mention-card"><img class="notion-link-mention-preview-thumbnail" src="https://imageio.forbes.com/specials-images/imageserve/688030c9ca6b13f36a68d152/0x0.jpg?format=jpg&amp;height=900&amp;width=1600&amp;fit=bounds" alt="What Is The San Francisco Consensus, Silicon Valley’s AI Prophecy?" referrerPolicy="same-origin"/><div class="notion-link-mention-preview-content"><p class="notion-link-mention-preview-title">What Is The San Francisco Consensus, Silicon Valley’s AI Prophecy?</p><p class="notion-link-mention-preview-description">The San Francisco Consensus predicts AI will revolutionize everything within 3-6 years. Visionary or hyperbole? Either way, it&#x27;s already redirecting trillions in capital.</p><div class="notion-link-mention-preview-footer"><img class="notion-link-mention-preview-icon" src="https://i.forbesimg.com/144X144-F.png" alt="Forbes" referrerPolicy="same-origin"/><span class="notion-link-mention-preview-provider">Forbes</span></div></div></article></div></span></div></main></div>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[OpenAI近期的资本操作令人眼花缭乱]]></title>
            <link>iqiujin.com/article/285bbc89-34a4-805d-a948-c1db2282f971</link>
            <guid>iqiujin.com/article/285bbc89-34a4-805d-a948-c1db2282f971</guid>
            <pubDate>Tue, 07 Oct 2025 00:00:00 GMT</pubDate>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-285bbc8934a4805da948c1db2282f971"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><div class="notion-text notion-block-285bbc8934a48066af43ded6f7fcdf4d">Sam Altman近期的一系列资本运作让人眼花缭乱，太牛了。当然他的牛气确实也是因为ChatGPT的产品确实够好，近期发布的产品和数据都让人对其前景有所期待。</div><div class="notion-text notion-block-285bbc8934a480868ccdc2f067ff698f">从 2023 年到今天，OpenAI 的生态规模出现了爆发式增长：</div><ul class="notion-list notion-list-disc notion-block-285bbc8934a480f9a686fd340569ec35"><li>每周活跃开发者从 200 万人 增长到 400 万人。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480b0a95afe41be77d159"><li>每周使用 ChatGPT 的用户则从 1 亿人 飙升至超过 8 亿人。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480509c9efdfde958d2fa"><li>有400万开发者在使用其API。</li></ul><div class="notion-text notion-block-285bbc8934a4800f92f0cbdb36c8b8a0">不仅如此，OpenAI API 每分钟处理的 Token 数量也从 2023 年的每分钟 3 亿个增长到今天的每分钟 60 亿个，增长了整整 20 倍。</div><div class="notion-blank notion-block-285bbc8934a480469b22de9ad695332a"> </div><blockquote class="notion-quote notion-block-285bbc8934a480b691d1d2e3401252ae"><div>英伟达向 OpenAI 投入 1000 亿元，OpenAI 用这笔资金购买英伟达的芯片并建设数据中心。实际上，英伟达并未直接出资，而是以市价 1000 亿元的芯片换取 OpenAI 10% 的股份。随后，OpenAI 将这些数据中心视为资产，向财团融资 500 亿元，用于向 AMD 采购芯片并再建一批数据中心。AMD 在获得这 500 亿元订单后，向 OpenAI 赠送了 10% 的认股权证，使其市值在短时间内从2000 亿元飙升至约 6000 亿元。此时，OpenAI 持有的 10% 股权价值约为 600 亿元，OpenAI 可以将这部分认购权转让给财团，偿还 500 亿元的贷款。最终，OpenAI 完成 IPO，二级市场的估值轻松突破万亿元。</div></blockquote><blockquote class="notion-quote notion-block-285bbc8934a48049b782e619c4229dc3"><div>英伟达凭借价值 1000 亿元的芯片，换回了二级市场上超过 1000 亿元的股票。</div><div class="notion-text notion-block-285bbc8934a48008b3f3e40569ab0bd6">结果如下：英伟达收回了等值 1000 亿元的芯片现金；</div><div class="notion-text notion-block-285bbc8934a4804491ede83ca6a2b07d">财团收回了 OpenAI 500 亿元的芯片抵押贷款；</div><div class="notion-text notion-block-285bbc8934a48050a18de81ff81fb2d2">AMD 获得了 500 亿元的芯片订单；</div><div class="notion-text notion-block-285bbc8934a4800095afea9603de2263">OpenAI 用 10% 的股权换取了市值约 1500 亿元的芯片，完成了数据中心的建设。</div><div class="notion-text notion-block-285bbc8934a480f897abd59a7a8cf51a">那么，这 1500 亿元的建设费用到底由谁承担？答案显而易见：最终由二级市场的投资者承担。</div></blockquote><div class="notion-blank notion-block-285bbc8934a480eebea6e4bf78b4b1ef"> </div><div class="notion-blank notion-block-285bbc8934a480a99c77eb8fc9faa281"> </div><div class="notion-text notion-block-285bbc8934a480e69847cc4912f1361b"><b>最终由二级市场的投资者承担。</b>但这并不是说二级市场的投资者是“冤大头”。可以这样理解：</div><ol start="1" class="notion-list notion-list-numbered notion-block-285bbc8934a4802aa807dbcafe62a5a6" style="list-style-type:decimal"><li><b>承担的是“成本”，换来的是“资产”</b>：二级市场投资者用1万亿（或更高）的现金，共同买下了整个OpenAI公司。这个公司此时的资产负债表上，确实拥有了价值1500亿的数据中心。投资者支付的价格，包含了这些已经建成的资产的成本。</li></ol><ol start="2" class="notion-list notion-list-numbered notion-block-285bbc8934a48083b702fc36f0251f7d" style="list-style-type:decimal"><li><b>支付的是“现在”，赌的是“未来”</b>：投资者之所以愿意用远超1500亿的价格（万亿估值）来购买股票，是因为他们相信OpenAI利用这些数据中心，在未来能够创造出远超万亿的价值（比如通过AI服务产生巨额利润）。</li></ol><ol start="3" class="notion-list notion-list-numbered notion-block-285bbc8934a480dd8e0ae28471f83c3e" style="list-style-type:decimal"><li><b>一个价值转移和放大的过程</b>：</li><ol class="notion-list notion-list-numbered notion-block-285bbc8934a480dd8e0ae28471f83c3e" style="list-style-type:lower-alpha"><ul class="notion-list notion-list-disc notion-block-285bbc8934a480e3889ad50af28d6c23"><li>NVIDIA和AMD付出了<b>产品（芯片）</b>。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a48033a645c181df785fc9"><li>财团付出了<b>短期现金（过桥贷款）</b>。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480689826c5235125b087"><li>OpenAI付出了<b>股权（未来的所有权）</b>。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a480cea5c3fb0cf1a7f1d7"><li>在这个过程中，通过金融操作和市场信心的放大，一个“需要1500亿现金才能建成”的项目，在几乎没有动用初始现金的情况下完成了。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4806d859df7a38601b621"><li>最后，二级市场的投资者作为一个整体，用<b>一笔巨额的现金</b>，将这个已经成型的、包含巨大潜力的“资产包”（整个OpenAI公司）买了下来。这笔钱，使得所有前期参与者（NVIDIA、财团等）都能成功退出并获利。</li></ul></ol></ol><div class="notion-text notion-block-285bbc8934a480abb501e0b6a607a975"><b>简单来说，这个模型就像一个接力赛：</b></div><ul class="notion-list notion-list-disc notion-block-285bbc8934a480e09556ccae5131b5a5"><li><b>第一棒（NVIDIA/AMD）</b>：用实物产品起跑。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4804884f1f2e49446553c"><li><b>第二棒（财团）</b>：用金融工具加速。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4805fad2ce587b7b23256"><li><b>第三棒（OpenAI）</b>：利用自身的潜力和市场预期，将雪球滚大。</li></ul><ul class="notion-list notion-list-disc notion-block-285bbc8934a4808990a2f0b1e7481e75"><li><b>终点线（二级市场投资者）</b>：用现金冲过终点，为整场比赛的价值“买单”，同时也拿到了比赛的“奖杯”（公司的所有权和未来的增长潜力）。</li></ul><div class="notion-blank notion-block-285bbc8934a480ad94d0db6297640dd7"> </div><div class="notion-callout notion-gray_background_co notion-block-285bbc8934a48021adf6dc4dffdb8f89"><div class="notion-page-icon-inline notion-page-icon-span"><span class="notion-page-icon" role="img" aria-label="💡">💡</span></div><div class="notion-callout-text"><div class="notion-text notion-block-285bbc8934a480289e75c3e18ddbae96">doubaoseed-1.6:商业活动中，“谁承担成本”取决于“谁放弃了资源并承担了风险”。英伟达用芯片换股权、财团提供贷款，本质是承担了OpenAI的早期风险，若公司成功，他们通过二级市场退出获利；若失败，风险由他们自己承担。二级市场只是风险和价值的“转移场所”，而非“成本的最终承担者”。</div></div></div><div class="notion-blank notion-block-285bbc8934a4808185c6fbcf16d453e5"> </div></main></div>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google Finance传统模式与AI模式对比]]></title>
            <link>iqiujin.com/article/284bbc89-34a4-80f4-a7f0-cde90af7e2f4</link>
            <guid>iqiujin.com/article/284bbc89-34a4-80f4-a7f0-cde90af7e2f4</guid>
            <pubDate>Tue, 07 Oct 2025 00:00:00 GMT</pubDate>
            <description><![CDATA[本文通过深入对比Google Finance传统版(2006-2017)与最新AI驱动Beta版的功能差异、用户体验和战略定位，揭示谷歌如何通过AI技术重构各个应用。]]></description>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-284bbc8934a480f4a7f0cde90af7e2f4"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><h2 class="notion-h notion-h1 notion-h-indent-0 notion-block-284bbc8934a480e7b137cefb81e612f1" data-id="284bbc8934a480e7b137cefb81e612f1"><span><div id="284bbc8934a480e7b137cefb81e612f1" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480e7b137cefb81e612f1" title="Google Finance的AI革命：传统版与Beta版深度对比分析"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">Google Finance的AI革命：传统版与Beta版深度对比分析</span></span></h2><div class="notion-text notion-block-284bbc8934a480138badffe7ba37db65">传统的财经频道</div><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-284bbc8934a4800cb0c8fdaffe5f2516"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A4a8c163c-6c63-43d0-9d76-ca1266656701%3Aimage.png?table=block&amp;id=284bbc89-34a4-800c-b0c8-fdaffe5f2516&amp;t=284bbc89-34a4-800c-b0c8-fdaffe5f2516" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-blank notion-block-284bbc8934a480578660d90ff73249d8"> </div><div class="notion-text notion-block-284bbc8934a48041854fc4bf8253bb04">AI修改后的财经频道，搜索框最为突出。搜索框代表chat和research的交互模式。</div><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-284bbc8934a48058ab89e9572027a36a"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3Ab8aa4e1d-f708-484c-af14-98ba71cdebc9%3Aimage.png?table=block&amp;id=284bbc89-34a4-8058-ab89-e9572027a36a&amp;t=284bbc89-34a4-8058-ab89-e9572027a36a" alt="notion image" loading="lazy" decoding="async"/></div></figure><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-284bbc8934a4801798b2dfcd687d7e4a"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A457410fa-0337-4cfc-ae50-a37ea854f146%3Aimage.png?table=block&amp;id=284bbc89-34a4-8017-98b2-dfcd687d7e4a&amp;t=284bbc89-34a4-8017-98b2-dfcd687d7e4a" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-blank notion-block-284bbc8934a48061910dc519e904c2dd"> </div><h3 class="notion-h notion-h2 notion-h-indent-1 notion-block-284bbc8934a480f9b0a9e5534496cf99" data-id="284bbc8934a480f9b0a9e5534496cf99"><span><div id="284bbc8934a480f9b0a9e5534496cf99" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480f9b0a9e5534496cf99" title="🎯 核心要点"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">🎯 核心要点</span></span></h3><div class="notion-text notion-block-284bbc8934a480ab87e5ebdf3aa94e15">本文通过深入对比Google Finance传统版(2006-2017)与最新AI驱动Beta版的功能差异、用户体验和战略定位，揭示谷歌如何通过AI技术重构各个应用。</div><h3 class="notion-h notion-h2 notion-h-indent-1 notion-block-284bbc8934a480e5a974e25f6a5f2bbf" data-id="284bbc8934a480e5a974e25f6a5f2bbf"><span><div id="284bbc8934a480e5a974e25f6a5f2bbf" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480e5a974e25f6a5f2bbf" title="📖 文章正文"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">📖 文章正文</span></span></h3><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a48024a84affe01ce7f299" data-id="284bbc8934a48024a84affe01ce7f299"><span><div id="284bbc8934a48024a84affe01ce7f299" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48024a84affe01ce7f299" title="一、Google Finance的战略摇摆：从数据门户到AI分析师"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">一、Google Finance的战略摇摆：从数据门户到AI分析师</span></span></h4><div class="notion-text notion-block-284bbc8934a48076a502f0ecc8622725">Google Finance的发展历程堪称谷歌产品战略演变的缩影：一个雄心勃勃的发布，一段时期的停滞，一次备受争议的转型，以及当前由AI驱动的重塑。</div><div class="notion-text notion-block-284bbc8934a48075873ec21a0abde40c"><b>2006年</b>：Google Finance正式上线，凭借简洁界面、无广告体验和创新功能（如新闻事件叠加图表）迅速成为金融信息服务领域的有力竞争者。</div><div class="notion-text notion-block-284bbc8934a48028bc51c59044df9668"><b>2017年</b>：一次灾难性的改版将其从功能丰富的独立平台转变为与谷歌搜索深度整合的简化工具，移除了投资组合、财务报表和股票筛选器等核心功能，引发专业用户强烈不满。</div><div class="notion-text notion-block-284bbc8934a480b99084f673ded2c79a"><b>2024-2025年</b>：面对用户持续批评和生成式AI的崛起，谷歌推出AI驱动的Beta版，试图利用其AI优势重新夺回市场地位。</div><div class="notion-text notion-block-284bbc8934a480acbdeac18bb8595944">这一演变轨迹揭示了谷歌产品战略中反复出现的模式：创新发布→最低维护→激进（可能失误的）重新设计→资源投入重建。Beta版不仅是功能更新，更是一次战略性的航向修正，试图纠正2017年&quot;低能化&quot;策略的错误，赢回高价值用户。</div><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a480dead98c062eb9a6ecb" data-id="284bbc8934a480dead98c062eb9a6ecb"><span><div id="284bbc8934a480dead98c062eb9a6ecb" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480dead98c062eb9a6ecb" title="二、传统版时代（2006-2017）：为成熟投资者打造的实用主义中心"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">二、传统版时代（2006-2017）：为成熟投资者打造的实用主义中心</span></span></h4><div class="notion-text notion-block-284bbc8934a4806fa788d1f4855a2ce1">传统版Google Finance培养了一批忠实用户，它不仅仅是数据查询工具，更是集研究、发现和监控于一体的综合性平台。</div><div class="notion-text notion-block-284bbc8934a480078d70fba50513b713"><b>核心优势功能</b>：</div><ol start="1" class="notion-list notion-list-numbered notion-block-284bbc8934a480f7ac1bee05a2a52200" style="list-style-type:decimal"><li><b>投资组合工具</b>：允许记录详细交易历史，实时计算投资组合总价值，与市场指数对比表现，并整合持仓股票的即将发生事件。这一功能将Google Finance从数据源提升为个性化投资管理仪表盘。</li></ol><ol start="2" class="notion-list notion-list-numbered notion-block-284bbc8934a4807f980eef7530475413" style="list-style-type:decimal"><li><b>全面财务数据</b>：提供完整的利润表、资产负债表和现金流量表，以及市盈率、市值、股息收益率等关键指标，为基本面分析提供坚实基础。</li></ol><ol start="3" class="notion-list notion-list-numbered notion-block-284bbc8934a480f5a450dc96bafa41a3" style="list-style-type:decimal"><li><b>集成分析工具</b>：基础但功能齐全的股票筛选器，支持按市值、市盈率等多维度筛选；先进的交互式图表，支持股票比较和新闻事件叠加；以及独特的Google国内趋势功能，利用搜索数据提供市场兴趣的领先指标。</li></ol><div class="notion-text notion-block-284bbc8934a48070b165d9f364aad3fa">传统版的界面简洁明了，优先考虑信息密度而非现代美学，深受需要快速获取全面数据的用户青睐。它创造了一个自成一体的生态系统，将发现、研究和监控流程集中在一处，带来极高的用户粘性。</div><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a48053b046e033104d8c02" data-id="284bbc8934a48053b046e033104d8c02"><span><div id="284bbc8934a48053b046e033104d8c02" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48053b046e033104d8c02" title="三、2017年大改版：备受争议的简洁化转向"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">三、2017年大改版：备受争议的简洁化转向</span></span></h4><div class="notion-text notion-block-284bbc8934a480bba81fd7803e6f4f02">2017年的全面改版是Google Finance历史上的分水岭，标志着从服务专业投资者向迎合普通用户的战略转变。</div><div class="notion-text notion-block-284bbc8934a480698023c69fa52e50f3"><b>功能&quot;墓地&quot;</b>：</div><ul class="notion-list notion-list-disc notion-block-284bbc8934a480c1929fc63ce90e43b3"><li>投资组合工具→极其简化的&quot;关注列表&quot;（无法记录购买价格和计算回报率）</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480a28ae8cb6597447435"><li>完整财务报表→仅保留利润表</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4801da9b2eb6c632ba03c"><li>股票筛选器→完全删除</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4800ca42bfb1db0c6ae40"><li>Google国内趋势→删除</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4805d819fd4235bb5d990"><li>新闻聚合→从广泛聚合转为仅展示少数主流发行商内容</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4807993c9f3625eb46fae"><li>详细公司信息→从路透社业务摘要转为简短维基百科描述</li></ul><div class="notion-text notion-block-284bbc8934a48084b6c7f517497a7b97">用户反应是压倒性的负面，将新界面描述为&quot;完全无用的垃圾&quot;、&quot;明显更差&quot;。甚至有专业交易员表示这是他们工作中遇到的&quot;最具破坏性的事情&quot;。</div><div class="notion-text notion-block-284bbc8934a4809382b6f4d980c84fbe">这次改版反映了&quot;整合优于功能&quot;的战略逻辑，将金融数据视为在主搜索引擎中呈现的一种内容类型，而非独立目的地产品。然而，这一决策错误地定位了目标受众，为休闲用户优化的同时，疏远了最投入、最有价值的用户群体。</div><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a48003b893c458b4e9cc7d" data-id="284bbc8934a48003b893c458b4e9cc7d"><span><div id="284bbc8934a48003b893c458b4e9cc7d" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48003b893c458b4e9cc7d" title="四、AI驱动的Beta版：谷歌对未来金融研究的愿景"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">四、AI驱动的Beta版：谷歌对未来金融研究的愿景</span></span></h4><div class="notion-text notion-block-284bbc8934a4800886abe272c46083b0">面对持续批评和AI时代到来，谷歌推出的Beta版不再是简单功能增补，而是基于AI的彻底重塑，围绕三大支柱：AI驱动的研究、高级图表和扩展数据。</div><div class="notion-text notion-block-284bbc8934a480a39e54ee5b8a88017c"><b>核心创新功能</b>：</div><div class="notion-text notion-block-284bbc8934a480bc9bdadc5dcfa0c712"><b>1 核心功能：对话式AI研究面板</b></div><ul class="notion-list notion-list-disc notion-block-284bbc8934a480a2af94e48d5dcc7efc"><li><b>自然语言查询</b>：这是Beta版最核心的创新。用户可以超越简单的股票代码搜索，用自然语言提出复杂的、比较性的问题，例如，“过去一年中，哪些科技股的表现超过了标普500指数？” 。AI能够理解问题并一次性综合信息，提供分析结果 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480118324fb4778cc2127"><li><b>情境理解</b>：该AI旨在理解金融术语，并能向不同专业水平的用户解释概念 。它提供的回答通常附有支持性链接和文档来源，增加了透明度 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480398564ddce38d4fa28"><li><b>限制与保障</b>：至关重要的是，该AI聊天机器人被明确设计为<em>不</em>提供直接的财务建议或股票推荐，这是规避监管风险的关键保障措施 。然而，信息不准确或产生“幻觉”的风险仍然是一个重大隐患 。</li></ul><div class="notion-text notion-block-284bbc8934a4809abb7bc1a9b6a71540"><b>2 高级图表与技术分析</b></div><div class="notion-text notion-block-284bbc8934a4808fa2baec69c1e3b4a0">这是对2017版过度简化图表的直接回应，旨在满足更专业用户的需求。</div><ul class="notion-list notion-list-disc notion-block-284bbc8934a4802bae7be83dad3b5d70"><li><b>新的图表类型</b>：Beta版引入了K线图（candlestick charts），这是技术分析的标准工具 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4803bacc3dddbbfc88b88"><li><b>技术指标</b>：用户现在可以应用移动平均包络线（moving average envelopes）等技术指标来跟踪长期趋势，并识别超买/超卖状况 。这使得该工具的功能更接近专业的交易平台。</li></ul><div class="notion-text notion-block-284bbc8934a4807f9d6ddcafdee2db90"><b>3 扩展的数据与实时新闻</b></div><ul class="notion-list notion-list-disc notion-block-284bbc8934a48023b7dbe9075b972ab5"><li><b>更广泛的资产覆盖</b>：平台现在包含更多关于大宗商品和更广泛加密货币的数据 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480dd8551cfbb6b5e0d32"><li><b>实时新闻流</b>：一个“即时”的实时新闻源被整合到界面中，允许用户在不离开页面的情况下跟踪影响市场的头条新闻 。AI还会根据用户的投资组合和兴趣策划这个新闻源，提供个性化体验 。</li></ul><div class="notion-text notion-block-284bbc8934a4802c83adff3c54dce2e1">AI聊天机器人的主要功能是将研究工作流程从“数据提取”重构为“问题解答”。这是一个根本性的用户交互转变。传统模式要求用户知道自己需要什么数据（如市盈率、现金流），然后逐个提取并自行综合。新模式允许用户从一个高层次的问题开始，由AI完成数据提取和初步的综合工作 。这降低了复杂研究的门槛，但也在用户和原始数据之间增加了一个抽象层和潜在的不确定性。</div><div class="notion-text notion-block-284bbc8934a4805eb41cf01abccc0504">同时，谷歌正战略性地将Beta版定位为其庞大数据索引的“AI前端”，充分利用其核心的搜索能力。平台不再像传统版那样是一个自给自足的数据库，而更像是一个连接整个金融网络的智能界面。AI生成的回答附带“来自全网的支持性文档”，这正是谷歌搜索的核心竞争力 。通过将此能力整合到Finance中，谷歌正在构建一个竞争对手难以复制的护城河。</div><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a480ea844eda03a41db4ea" data-id="284bbc8934a480ea844eda03a41db4ea"><span><div id="284bbc8934a480ea844eda03a41db4ea" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480ea844eda03a41db4ea" title="五、传统版与Beta版的正面交锋：功能矩阵对比"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">五、传统版与Beta版的正面交锋：功能矩阵对比</span></span></h4><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a4806fbd52e02c73ac46de" data-id="284bbc8934a4806fbd52e02c73ac46de"><span><div id="284bbc8934a4806fbd52e02c73ac46de" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a4806fbd52e02c73ac46de" title="核心功能对比分析"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">核心功能对比分析</span></span></h4><div class="notion-blank notion-block-284bbc8934a480cb8067ef831b613f4e"> </div><table class="notion-simple-table notion-block-284bbc8934a48044adc2c0ff4af0e5c1"><tbody><tr class="notion-simple-table-row notion-block-284bbc8934a4809285c7c956f93a9447"><td class="" style="width:120px"><div class="notion-simple-table-cell">功能</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">传统版 (2017年前)</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">AI驱动的Beta版</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">转变分析</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4804381b4c9ca01048428"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>投资组合管理</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">提供全面的跟踪功能，包括购买日期、价格和股数；计算投资组合价值和总回报；与指数进行业绩比较   。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">在被移除后重新引入，功能基本恢复，支持输入购买数据和计算回报   。增加了“模拟”模式   。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">经历了一次灾难性的移除后，核心功能已基本恢复到传统版水平，但拥有更现代的界面。2017年至2020年间的功能缺失是一个重大的战略失误。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a48071825ac93fce7da823"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>财务报表</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">提供利润表、资产负债表和现金流量表  。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">界面主要突出利润表数据。对资产负债表和现金流量表的直接访问并非突出功能，尽管数据可能通过API获取。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">一次显著的降级。Beta版优先考虑AI综合的见解，而非对原始、详细财务报表的直接访问，这对基本面分析师来说是一个重大损失。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a480498803c849917a1c88"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>股票筛选器</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">包含一个基础但功能齐全的股票筛选器，用于根据基本面指标进行发现 。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">没有内置的股票筛选器。被AI研究面板的自然语言查询功能所取代  。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">从用户驱动的过滤转变为AI驱动的查询。这为复杂问题提供了更大的灵活性，但牺牲了传统筛选器的结构化、系统性方法。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a48092954af0cc7d3174c4"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>图表与分析</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">交互式图表，支持新闻叠加和基本的股票比较。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">高级图表，支持K线图和技术指标（如移动平均包络线）。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">一次清晰而重大的升级。Beta版提供了更适用于技术分析的工具，解决了所有先前版本的一个主要弱点。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4800e8163d7e2f2ee7919"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>独特数据功能</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">Google国内趋势（基于搜索量数据）。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">AI驱动的对话式研究面板和个性化新闻源。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">两种独特优势的取舍。传统版提供了独特的<em>原始数据</em>（搜索趋势），而Beta版提供了独特的<em>数据交互方式</em>（AI对话）。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4805aa61bc4ef47637d0f"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>数据覆盖范围</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">主要覆盖股票、指数、货币。通过Google News提供广泛新闻 。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">扩展了对大宗商品和加密货币的覆盖。提供实时、策划的新闻源。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">顺应现代市场兴趣的扩展。从广泛新闻到策划新闻的转变，与个性化主题相符。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4805396adc0ce2c35df35"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>生态系统整合</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">与Google News有基本整合。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">与谷歌搜索、AI（Gemini）和Google Sheets（<code class="notion-inline-code">GOOGLEFINANCE</code>函数）深度整合 。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">Beta版从根本上是更广泛的谷歌生态系统的一个组成部分，而传统版更像一个独立产品。</div></td></tr></tbody></table><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a480ec8cabf767682bb0e0" data-id="284bbc8934a480ec8cabf767682bb0e0"><span><div id="284bbc8934a480ec8cabf767682bb0e0" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480ec8cabf767682bb0e0" title="关键维度对比："><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">关键维度对比：</span></span></h4><ol start="1" class="notion-list notion-list-numbered notion-block-284bbc8934a480fb9304cd28ca20a436" style="list-style-type:decimal"><li><b>数据深度与可访问性</b>：</li><ol class="notion-list notion-list-numbered notion-block-284bbc8934a480fb9304cd28ca20a436" style="list-style-type:lower-alpha"><ul class="notion-list notion-list-disc notion-block-284bbc8934a4800e8dd7c79aaf899e04"><li>传统版：优先原始数据直接访问，用户可自行分析解读</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480b3bca6d304fccde5b2"><li>Beta版：优先AI综合答案，更快但可能掩盖底层数据并引入不准确风险</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48097aae6fab8b449e69a"><li>转变反映了&quot;数据民主化&quot;与&quot;专家赋能&quot;的权衡</li></ul></ol></ol><ol start="2" class="notion-list notion-list-numbered notion-block-284bbc8934a480aab906dad9818383e0" style="list-style-type:decimal"><li><b>用户主导与发现</b>：</li><ol class="notion-list notion-list-numbered notion-block-284bbc8934a480aab906dad9818383e0" style="list-style-type:lower-alpha"><ul class="notion-list notion-list-disc notion-block-284bbc8934a480439bf1d7f4692205c9"><li>传统版：用户通过筛选器完全控制，支持系统性探索</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480828f98ffafe71ded34"><li>Beta版：依赖用户清晰表述需求，适合特定问题但不适合广泛探索</li></ul></ol></ol><ol start="3" class="notion-list notion-list-numbered notion-block-284bbc8934a48027be04e453705da60a" style="list-style-type:decimal"><li><b>共同缺陷</b>：</li><ol class="notion-list notion-list-numbered notion-block-284bbc8934a48027be04e453705da60a" style="list-style-type:lower-alpha"><ul class="notion-list notion-list-disc notion-block-284bbc8934a4808298d9d4b6f00dee30"><li>两者都未提供原生集成的期权链数据</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480e691e9fa78601fe373"><li>缺乏主流提供商的ESG评级数据</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48091b34feaeb78f1888d"><li>这些缺口使Google Finance在某些专业领域不够完整</li></ul></ol></ol><h4 class="notion-h notion-h3 notion-h-indent-2 notion-block-284bbc8934a480cc944cd19e38cd4d0c" data-id="284bbc8934a480cc944cd19e38cd4d0c"><span><div id="284bbc8934a480cc944cd19e38cd4d0c" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480cc944cd19e38cd4d0c" title="六、Beta版的AI优势与局限"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">六、Beta版的AI优势与局限</span></span></h4><div class="notion-text notion-block-284bbc8934a48020b5b2db6532d30eac">Beta版最大的创新在于其AI驱动的对话式研究面板，这代表了金融数据交互方式的根本转变：</div><div class="notion-text notion-block-284bbc8934a48040a24afb4c82c2fb9a"><b>AI优势</b>：</div><ul class="notion-list notion-list-disc notion-block-284bbc8934a480da8d06c4d0726d24d5"><li>自然语言处理使复杂查询成为可能，降低研究门槛</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4801cabcfc7a35f63b498"><li>情境理解能力，可向不同专业水平用户解释金融概念</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480faa63af710dec1e4fc"><li>个性化新闻和见解，减少信息过载</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480309fd6d0088a9f9e02"><li>与谷歌搜索深度整合，提供支持性链接和文档来源</li></ul><div class="notion-text notion-block-284bbc8934a4803b92c7e0aecfe4d510"><b>AI局限与风险</b>：</div><ul class="notion-list notion-list-disc notion-block-284bbc8934a48082a5fdc48aa0cc5ac3"><li>不提供直接财务建议，功能受限</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4802ab8b1c6ffd0d1cdb7"><li>信息不准确或&quot;幻觉&quot;的风险</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480bb8a03e610abcb003b"><li>在用户与原始数据间增加抽象层，可能掩盖关键细节</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4806eb7f4fceebd739036"><li>依赖用户提问质量，&quot;问不出好问题就得不到好答案&quot;</li></ul><div class="notion-text notion-block-284bbc8934a4803f99abeb8d16ab8767">Beta版的AI功能本质上是将研究工作流程从&quot;我需要什么数据？如何获取？如何分析？&quot;转变为&quot;我想知道什么？&quot;，这对投资研究的普及化有积极意义，但可能无法完全满足专业投资者的深度需求。</div><h3 class="notion-h notion-h2 notion-h-indent-1 notion-block-284bbc8934a48062a19ed136848583d6" data-id="284bbc8934a48062a19ed136848583d6"><span><div id="284bbc8934a48062a19ed136848583d6" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48062a19ed136848583d6" title="💡 关键洞察"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">💡 关键洞察</span></span></h3><ol start="1" class="notion-list notion-list-numbered notion-block-284bbc8934a48074b61ffdecd9a2a081" style="list-style-type:decimal"><li>Google Finance的产品战略始终与谷歌整体企业战略保持一致，从独立产品到搜索整合，再到如今的AI整合，反映了科技巨头资源分配的优先级变化。</li></ol><ol start="2" class="notion-list notion-list-numbered notion-block-284bbc8934a480c6ac64f4bfaeff3067" style="list-style-type:decimal"><li>2017年改版是典型的&quot;过度简化&quot;错误案例，通过削减专业功能迎合大众市场，最终疏远了最有价值的用户群体，这一教训对所有产品管理者都具有警示意义。</li></ol><ol start="3" class="notion-list notion-list-numbered notion-block-284bbc8934a480cf8427dd458c1759d0" style="list-style-type:decimal"><li>Beta版的AI整合代表了金融信息服务的未来趋势，但不应以牺牲数据深度为代价，理想状态是AI辅助而非替代深度分析能力。</li></ol><ol start="4" class="notion-list notion-list-numbered notion-block-284bbc8934a480ab95f0cc4ed9ff455e" style="list-style-type:decimal"><li>产品设计中&quot;数据民主化&quot;与&quot;专家赋能&quot;可以共存，最佳解决方案是为不同专业水平用户提供分层功能体验，而非一刀切。</li></ol><ol start="5" class="notion-list notion-list-numbered notion-block-284bbc8934a4802d87aefa385301adde" style="list-style-type:decimal"><li>谷歌生态系统整合是Beta版的隐形优势，特别是与Google Sheets的联动，可能成为专业用户弥补功能缺口的途径。</li></ol><ol start="6" class="notion-list notion-list-numbered notion-block-284bbc8934a48083beebc780e1cdf986" style="list-style-type:decimal"><li>金融数据服务的竞争将越来越集中在AI理解能力和数据交互体验上，而不仅仅是数据本身的覆盖广度。</li></ol><h3 class="notion-h notion-h2 notion-h-indent-1 notion-block-284bbc8934a480cab840d79461a96771" data-id="284bbc8934a480cab840d79461a96771"><span><div id="284bbc8934a480cab840d79461a96771" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480cab840d79461a96771" title="🏷️ 标签"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">🏷️ 标签</span></span></h3><div class="notion-text notion-block-284bbc8934a4807bbd44c798cd576902"><code class="notion-inline-code">Google Finance</code> <code class="notion-inline-code">金融科技</code> <code class="notion-inline-code">AI金融</code> <code class="notion-inline-code">投资工具</code> <code class="notion-inline-code">产品对比</code> <code class="notion-inline-code">金融数据分析</code> <code class="notion-inline-code">技术分析</code> <code class="notion-inline-code">基本面分析</code> <code class="notion-inline-code">谷歌产品战略</code> <code class="notion-inline-code">金融AI应用</code></div><h3 class="notion-h notion-h2 notion-h-indent-1 notion-block-284bbc8934a480bb88a7c8cc8b30562c" data-id="284bbc8934a480bb88a7c8cc8b30562c"><span><div id="284bbc8934a480bb88a7c8cc8b30562c" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480bb88a7c8cc8b30562c" title="📊 内容评估"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">📊 内容评估</span></span></h3><ul class="notion-list notion-list-disc notion-block-284bbc8934a48079b966e96106b8ff59"><li><b>技术深度</b>: ⭐⭐⭐⭐ (深入分析了产品功能差异和技术实现)</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480a38468da8f1c5280b6"><li><b>实用性</b>: ⭐⭐⭐⭐⭐ (提供了详细对比和具体行动建议)</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480d49debe40f88a44e0c"><li><b>可读性</b>: ⭐⭐⭐⭐ (结构清晰，案例丰富，平衡了技术与非技术内容)</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4807fbbf8c401968d3794"><li><b>独特性</b>: ⭐⭐⭐⭐ (基于详细的产品历史和功能分析，提供了独到见解)</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480278ed3f8ccef6e3681"><li><b>改进建议</b>: 可增加更多实际操作截图对比，以及对Beta版AI准确性的测试数据，进一步增强说服力。</li></ul></main></div>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google AI mode的研究]]></title>
            <link>iqiujin.com/article/284bbc89-34a4-80c7-993b-d54795788df0</link>
            <guid>iqiujin.com/article/284bbc89-34a4-80c7-993b-d54795788df0</guid>
            <pubDate>Tue, 07 Oct 2025 00:00:00 GMT</pubDate>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-284bbc8934a480c7993bd54795788df0"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><div class="notion-blank notion-block-284bbc8934a4803c88c2da9406cfec2a"> </div><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-284bbc8934a480a9b3b2de2d58053556"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A7225cda1-b4c1-4733-91af-ee369843e397%3Aimage.png?table=block&amp;id=284bbc89-34a4-80a9-b3b2-de2d58053556&amp;t=284bbc89-34a4-80a9-b3b2-de2d58053556" alt="notion image" loading="lazy" decoding="async"/></div></figure><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-284bbc8934a480e89751e0ee21d3e587"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3Ade7dcf8f-b3ab-410f-bde1-e3c08b2008c9%3Aimage.png?table=block&amp;id=284bbc89-34a4-80e8-9751-e0ee21d3e587&amp;t=284bbc89-34a4-80e8-9751-e0ee21d3e587" alt="notion image" loading="lazy" decoding="async"/></div></figure><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-284bbc8934a4806d8f73e183b790e18b" data-id="284bbc8934a4806d8f73e183b790e18b"><span><div id="284bbc8934a4806d8f73e183b790e18b" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a4806d8f73e183b790e18b" title="一、Google自己的宣传"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">一、Google自己的宣传</span></span></h3><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a48003a6aad21354466b37" data-id="284bbc8934a48003a6aad21354466b37"><span><div id="284bbc8934a48003a6aad21354466b37" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48003a6aad21354466b37" title="Ask anything-融入了知乎的问题模式"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">Ask anything-融入了知乎的问题模式</span></span></h4><div class="notion-text notion-block-284bbc8934a480a9a606ed7aa8d54a26">Trying to wrap your head around a new topic? Looking for specific recommendations? Need help deciding between two products? Just ask.</div><div class="notion-blank notion-block-284bbc8934a480b883d6f8483ced148a"> </div><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480a9bee0fc5682b7163a" data-id="284bbc8934a480a9bee0fc5682b7163a"><span><div id="284bbc8934a480a9bee0fc5682b7163a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480a9bee0fc5682b7163a" title="Ask any way-这种模式豆包的呼叫模式是创新"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">Ask any way-这种模式豆包的呼叫模式是创新</span></span></h4><div class="notion-text notion-block-284bbc8934a480399b12f6257e5a1d54"><span class="notion-blue"><b>Type, talk, snap a photo, or upload an image</b></span><b>.</b> AI Mode uses advanced multi-modal capabilities to deeply understand your question, no matter how you ask.</div><div class="notion-blank notion-block-284bbc8934a480899577e55244b2185b"> </div><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a4802797d6dec6c949acae" data-id="284bbc8934a4802797d6dec6c949acae"><span><div id="284bbc8934a4802797d6dec6c949acae" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a4802797d6dec6c949acae" title="Research made easy"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title"><b>Research made easy</b></span></span></h4><div class="notion-text notion-block-284bbc8934a480668f05c38185b852e1"><b>
Dive into any topic</b>
Ask your question all in one go, with all of the details you care about. AI Mode intelligently organizes information and answers your question in a simple and intuitive way, with links to explore more on the web.<b>
Go even deeper with Deep Search</b>
For questions where you want an even more thorough response, try Deep Search in AI Mode — our most advanced research tool in Google Search. Built with the Gemini 2.5 Pro model, it browses hundreds of sites and reasons across them, crafting a comprehensive, fully-cited report in just minutes. Deep Search is available to Google AI subscribers in Labs.<b>
Keep asking,keep discovering </b>Ask follow-up questions to dig deeper or change course. Easily revisit past searches to pick up where you left off.<b>
Reliable results from the best of the web </b>AI Mode connects you to high quality information from the best of the web. You’ll find helpful links to evaluate sources and explore further from a wide range of perspectives.</div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-284bbc8934a48095b7e0ea643166cea0" data-id="284bbc8934a48095b7e0ea643166cea0"><span><div id="284bbc8934a48095b7e0ea643166cea0" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a48095b7e0ea643166cea0" title="二、范式转移：谷歌AI模式深度解析"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">二、范式转移：谷歌AI模式深度解析</span></span></h3><div class="notion-text notion-block-284bbc8934a480a69d73c12798a08cc0">谷歌AI模式的出现，代表了一种技术物种的跃迁。它不再是对现有搜索框架的修补，而是基于全新的交互模型、核心能力和技术引擎的彻底重构，旨在将搜索引擎从信息检索工具转变为知识综合与推理的伙伴。</div><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a4800ab45ad5dd4081ef4b" data-id="284bbc8934a4800ab45ad5dd4081ef4b"><span><div id="284bbc8934a4800ab45ad5dd4081ef4b" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a4800ab45ad5dd4081ef4b" title="2.1 从关键词到对话：全新的用户交互模型"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">2.1 从关键词到对话：全新的用户交互模型</span></span></h4><ul class="notion-list notion-list-disc notion-block-284bbc8934a4802d8fbee623bcf8e950"><li><b>自然语言交互：</b> AI模式的核心是让用户能够以自然、流畅的对话方式提出问题，彻底摆脱了过去那种生硬、片段化的关键词查询模式 。用户现在可以一次性输入包含多个约束和细节的复杂问题，例如“帮我找一个适合四个人、周五晚上七点、在市中心、安靜的意大利餐厅” 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480409ab4e456212e622d"><li><b>有上下文的多轮对话：</b> 传统搜索的每次查询都是孤立的，而AI模式则能在连续的问答中保持上下文记忆 。用户可以对AI生成的答案进行追问，比如在得到餐厅推荐后继续问“哪家有户外座位？”或者“另一家的评价如何？” 。这种交互方式更接近于与人类专家对话，允许用户进行探索式、层层递进的信息发掘。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480e2a3c0c26d09e4538e" data-id="284bbc8934a480e2a3c0c26d09e4538e"><span><div id="284bbc8934a480e2a3c0c26d09e4538e" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480e2a3c0c26d09e4538e" title="2.2 核心能力：多模态、代理化与超个性化"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">2.2 核心能力：多模态、代理化与超个性化</span></span></h4><ul class="notion-list notion-list-disc notion-block-284bbc8934a480de934cd2311fb8dd04"><li><b>多模态（“Ask Any Way”）：</b> AI模式深度整合了多模态输入能力，用户不仅可以通过文字，还可以通过语音、拍摄照片或上传图片来发起查询 。这项能力植根于Google Lens等现有技术，但通过生成式AI的加持，能够对图像内容进行更深层次的理解并提供富有情境的综合性回答 。例如，用户可以拍下一株植物的照片，然后提问“这种植物需要多久浇一次水？”。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48074a01fdc58c00b6565"><li><b>代理化能力（“Getting Things Done”）：</b> AI模式正在演变为一个能够代表用户执行任务的智能代理（agent）。它不仅仅是提供信息，更致力于帮助用户完成现实世界中的任务。目前，这项能力已开始应用于餐厅预订等场景。AI模式可以根据用户的复杂需求，跨多个预订平台和网站搜索实时可用的席位，然后直接提供预订链接，极大地简化了用户的决策和操作流程 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a480c68bb4fa3f22031b82"><li><b>超个性化：</b> AI模式将个性化提升到了一个新的水平。它不仅利用用户的地理位置和语言等基本信息，还会结合用户的历史对话、在谷歌搜索和地图中的活动记录来推断其深层偏好 。例如，如果系统推断出用户偏爱素食和意大利菜，那么在推荐午餐地点时，会优先展示符合这些偏好的选项 。这使得搜索结果不再是千人一面，而是为每个用户量身定制的建议。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480bcb3bcfcd6929766da" data-id="284bbc8934a480bcb3bcfcd6929766da"><span><div id="284bbc8934a480bcb3bcfcd6929766da" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480bcb3bcfcd6929766da" title="2.3 技术引擎室：Gemini 2.5 Pro、查询扇出与深度搜索"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">2.3 技术引擎室：Gemini 2.5 Pro、查询扇出与深度搜索</span></span></h4><div class="notion-text notion-block-284bbc8934a48064bdf6e97deaa5d9b6">AI模式的强大能力背后，是谷歌最前沿的人工智能技术在支撑。</div><ul class="notion-list notion-list-disc notion-block-284bbc8934a480efa327dc356c8bb7aa"><li><b>Gemini 2.5 Pro——推理核心：</b> AI模式由谷歌最先进的Gemini 2.5 Pro模型驱动。该模型被描述为一个“会思考的模型”（thinking model），在推理、编码和多模态理解方面达到了业界顶尖水平 。其最引人注目的特性之一是拥有高达100万个令牌（token）的超长上下文窗口，这使其能够同时理解和分析海量的文本、代码库、图片、音频和视频数据，为深度综合信息提供了前所未有的能力 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48094a57ddeec3fbc4a3a"><li><b>“查询扇出”（Query Fan-Out）——信息收集机制：</b> 这是AI模式在架构上的一项关键创新。当接收到用户的复杂查询时，系统不会只执行一次搜索，而是会智能地将其分解为多个相关的子主题，并针对每个子主题同时发起并行的搜索，这些搜索会跨越多个数据源 。这种“扇出”机制使得AI模式能够在短时间内收集到比传统单一搜索更广泛、更深入的信息。随后，Gemini模型会将这些来自四面八方的信息碎片进行推理和整合，最终生成一个逻辑连贯、内容丰富的综合性答案 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4803f8612c552105b9293"><li><b>深度搜索（Deep Search）——终极综合引擎：</b> 对于那些需要进行详尽研究的超复杂问题，谷歌为订阅用户提供了“深度搜索”功能。在这一模式下，Gemini 2.5 Pro会自主地浏览数百个网站，在不同来源的信息之间进行交叉比对和推理，识别出核心主题、观点差异甚至信息矛盾之处，并在几分钟内生成一份结构完整、引用齐全的深度研究报告 。这代表了该系统信息综合能力的顶峰，它解决了多步规划和长时间推理等重大的技术挑战，将过去需要数小时甚至数天的人工研究工作自动化 。</li></ul><div class="notion-text notion-block-284bbc8934a480e18dcdc3c3939d6f69">这一系列的技术革新，共同促成了一次架构上的倒置。在传统搜索中，是用户从搜索引擎提供的链接列表中“拉取”（pull）信息。而在AI模式中，是AI从海量信源中主动“综合”（synthesize）信息，然后将一个全新的、经过加工的知识产物“推送”（push）给用户。在这个新范式中，信息的基本单位不再是承载信息的<em>文档</em>，而是信息本身——可以从任何来源被提取、验证和重组的<em>事实</em>或<em>概念</em>。用户的角色也随之改变，从一个需要在信息海洋中筛选、评估信源的<em>研究者</em>，转变为一个接收预消化知识摘要的<em>受益者</em>。</div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-284bbc8934a480eba079cbdb04190815" data-id="284bbc8934a480eba079cbdb04190815"><span><div id="284bbc8934a480eba079cbdb04190815" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480eba079cbdb04190815" title="三：对比分析架构与功能"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">三：对比分析架构与功能</span></span></h3><div class="notion-text notion-block-284bbc8934a4807bae75f3b45b203ed4">为了更清晰地揭示谷歌AI模式所带来的突破，本节将对新旧两种范式在核心功能层面进行直接的并列比较，从而凸显它们之间存在的巨大鸿沟。</div><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480889a16cedd93279183" data-id="284bbc8934a480889a16cedd93279183"><span><div id="284bbc8934a480889a16cedd93279183" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480889a16cedd93279183" title="3.1 查询解读：语义意图 vs. 词法匹配"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">3.1 查询解读：语义意图 vs. 词法匹配</span></span></h4><ul class="notion-list notion-list-disc notion-block-284bbc8934a48010bde1ccb9649e1cf7"><li><b>传统搜索：</b> 传统搜索引擎的查询解读主要依赖于关键词匹配。尽管后期引入了像RankBrain这样的人工智能系统来推断词语间的概念关系，但其根本逻辑仍然是寻找那些在文本内容上与查询词条或相关概念相匹配的文档 。它的目标是回答“哪些文档包含了这些词？”。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a4803d827cfcfeb5aa60ce"><li><b>AI模式：</b> AI模式则运用了先进的自然语言理解（NLU）技术，旨在洞察用户查询背后深层的<em>意图</em>（intent）、上下文和细微差别，即便这些意图没有被明确表述出来 。它能理解像“这里有点冷”这样的日常用语可能隐含着一个关闭窗户或调高暖气的请求，而不仅仅是搜索“冷”这个词的定义 。这种能力使其能够回答用户</li><ul class="notion-list notion-list-disc notion-block-284bbc8934a4803d827cfcfeb5aa60ce"><div class="notion-text notion-block-284bbc8934a4803a9f4cf755153d7254"><em>真正想问</em>的问题，而不仅仅是他们<em>输入</em>的那个问题。</div></ul></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480b99e0bfac0f62a8305" data-id="284bbc8934a480b99e0bfac0f62a8305"><span><div id="284bbc8934a480b99e0bfac0f62a8305" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480b99e0bfac0f62a8305" title="3.2 信息呈现：链接列表 vs. 综合答案"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">3.2 信息呈现：链接列表 vs. 综合答案</span></span></h4><ul class="notion-list notion-list-disc notion-block-284bbc8934a4808d8c65de4ef4712e43"><li><b>传统搜索：</b> 搜索结果页（SERP）的核心产物是一个由“十个蓝色链接”组成的排名列表。这些链接是指向外部文档的路标，用户的探索之旅从点击这些链接开始 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48004bcf0d8f2c1b96d1c"><li><b>AI模式：</b> 其主要输出是一个由AI实时生成的摘要，即“AI概览”（AI Overview）。这个摘要直接回答了用户的问题，并且通常以段落、项目符号列表、对比表格甚至交互式图表等富媒体形式呈现 。用户的探索之旅很可能在这个结果页上就已经</li><ul class="notion-list notion-list-disc notion-block-284bbc8934a48004bcf0d8f2c1b96d1c"><div class="notion-text notion-block-284bbc8934a48054b3b1e91526553c62"><em>结束</em>，这直接导致了“零点击搜索”（zero-click search）现象的出现 。</div></ul></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480d49a61d1144cd9a489" data-id="284bbc8934a480d49a61d1144cd9a489"><span><div id="284bbc8934a480d49a61d1144cd9a489" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480d49a61d1144cd9a489" title="3.3 数据粒度：页面级权威 vs. 段落级相关性"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">3.3 数据粒度：页面级权威 vs. 段落级相关性</span></span></h4><ul class="notion-list notion-list-disc notion-block-284bbc8934a480a287e0db2b8f867d33"><li><b>传统搜索：</b> 像PageRank这样的排名系统主要在页面或网站级别上运作。它们通过分析链接、域名历史等宏观信号来评估一个文档的整体权威性 。</li></ul><ul class="notion-list notion-list-disc notion-block-284bbc8934a48031bd8bd90b0f7fb3b3"><li><b>AI模式：</b> AI模式则利用了段落排名（passage ranking）等技术，能够识别并从一个网页中提取出与特定子查询最相关的单个段落或句子 。这意味着系统可以从一个权威新闻网站中提取一个定义，再从一个专业论坛中提取一个用户经验，将它们融合在同一个答案中。这种做法从根本上解构了“单个页面排名”的概念，将评估的粒度从整个文档细化到了信息本身。</li></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-284bbc8934a480e68e02e765c1250e12" data-id="284bbc8934a480e68e02e765c1250e12"><span><div id="284bbc8934a480e68e02e765c1250e12" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a480e68e02e765c1250e12" title="表1：传统搜索引擎 vs. 谷歌AI模式功能对比"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">表1：传统搜索引擎 vs. 谷歌AI模式功能对比</span></span></h4><table class="notion-simple-table notion-block-284bbc8934a48019852afe3c4cc8e636"><tbody><tr class="notion-simple-table-row notion-block-284bbc8934a480929235dd5f40286e1f"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>属性</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>传统搜索引擎</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>谷歌AI模式</b></div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4805ab822da4e6003c4b4"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>用户输入</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">基于关键词的查询，通常较为零散。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">对话式、自然语言查询；多模态（文本、语音、图像）。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a48070803cc71f800358d6"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>核心逻辑</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">词法匹配与文档检索。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">语义理解与信息综合。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a480c0abaec449b80fcc69"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>基本单位</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">网页/文档。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">事实/概念/段落。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a4805d9c76c123ccf2342f"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>输出格式</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">排名链接列表（SERP）。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">生成式摘要（AI概览）、报告、交互元素。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a48022add3e4bcde11e6f3"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>交互模型</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">单次、无状态的查询。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">多轮、有状态的对话。</div></td></tr><tr class="notion-simple-table-row notion-block-284bbc8934a480cd8f9bf410dd0f62d1"><td class="" style="width:120px"><div class="notion-simple-table-cell"><b>核心目标</b></div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">找到最权威的<b>信源</b>。</div></td><td class="" style="width:120px"><div class="notion-simple-table-cell">提供最全面的<b>答案</b>。</div></td></tr></tbody></table><div class="notion-blank notion-block-284bbc8934a480d69f5fd1a4b3cf6ddb"> </div><h2 class="notion-h notion-h1 notion-h-indent-0 notion-block-284bbc8934a4801e9ba8dad02239af5a" data-id="284bbc8934a4801e9ba8dad02239af5a"><span><div id="284bbc8934a4801e9ba8dad02239af5a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#284bbc8934a4801e9ba8dad02239af5a" title="来源："><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">来源：</span></span></h2><ol start="1" class="notion-list notion-list-numbered notion-block-284bbc8934a48066b86dc94d18de0baf" style="list-style-type:decimal"><li>google</li></ol><ol start="2" class="notion-list notion-list-numbered notion-block-284bbc8934a48030b566d7f1f9ad69b1" style="list-style-type:decimal"><li>gemini deepresearch</li></ol></main></div>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI处理长任务能力每7个月翻一倍]]></title>
            <link>iqiujin.com/article/27cbbc89-34a4-8075-98d3-eb42d2dc6cce</link>
            <guid>iqiujin.com/article/27cbbc89-34a4-8075-98d3-eb42d2dc6cce</guid>
            <pubDate>Sun, 28 Sep 2025 00:00:00 GMT</pubDate>
            <description><![CDATA[AI在多方面已经超过人类平均水平，在短任务处理上成功率很高，在长任务上也显现出“每7个月能力翻一倍”的快速发展趋势，针对这个情况，我们需要重点关注和做成应对。]]></description>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-27cbbc8934a4807598d3eb42d2dc6cce"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-27cbbc8934a480a5b65fdd46e86be81a" data-id="27cbbc8934a480a5b65fdd46e86be81a"><span><div id="27cbbc8934a480a5b65fdd46e86be81a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#27cbbc8934a480a5b65fdd46e86be81a" title="AI已经在多方面超越人类平均水平"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">AI已经在多方面超越人类平均水平</span></span></h3><div class="notion-text notion-block-27cbbc8934a4802f9141f9bb793fc504">当前前沿人工智能在文本预测和知识任务方面远超人类。它们在大多数考试类问题上的表现甚至优于专家，而成本却仅为人类的一小部分。通过针对特定任务进行一些调整，它们还可以成为许多应用中的实用工具。</div><ul class="notion-list notion-list-disc notion-block-27cbbc8934a48029b943efd8512a7832"><li><b>图像识别 (Image recognition)</b>: 约在2016年超越人类平均水平。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a4808cb0daf7c8f9122c38"><li><b>语音识别 (Speech recognition)</b>: 约在2017年超越人类平均水平。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480f890a5d99954b24f36"><li><b>手写识别 (Handwriting recognition)</b>: 约在2017-2018年超越人类平均水平。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a4802e885ad8d5f0ef2493"><li><b>阅读理解 (Reading comprehension)</b>: 约在2018年超越人类平均水平。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a4808bb607cf1e09ba49e7"><li><b>语言理解 (Language understanding)</b>: 约在2020年超越人类平均水平。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a48027ae8bcda77ef2cd25"><li><em>“预测性推理” (Predictive reasoning) 还是在接近人类平均水平中。</em></li></ul><div class="notion-text notion-block-27cbbc8934a48091af07e35f67c1edf7">然而，即使是最优秀的人工智能代理，目前也无法独立完成实质性项目，也无法直接取代人类劳动力。它们甚至无法可靠地处理相对低技能、基于计算机的工作。</div><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-27cbbc8934a480eeadbcd0f042eb40d7"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A66466b28-e27c-43d5-90f6-abd152a7c6a3%3Aimage.png?table=block&amp;id=27cbbc89-34a4-80ee-adbc-d0f042eb40d7&amp;t=27cbbc89-34a4-80ee-adbc-d0f042eb40d7" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-blank notion-block-27cbbc8934a480ba9047e4bf50c2883a"> </div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-27cbbc8934a48033be7be603809fba14" data-id="27cbbc8934a48033be7be603809fba14"><span><div id="27cbbc8934a48033be7be603809fba14" class="notion-header-anchor"></div><a class="notion-hash-link" href="#27cbbc8934a48033be7be603809fba14" title="METR专门对人工智能处理长时间的任务进行了分析研究。"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">METR专门对人工智能处理长时间的任务进行了分析研究。</span></span></h3><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-27cbbc8934a480688d2fd121dfda93fc"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A236ab9e7-7f7b-432b-9efb-de5a2890a288%3Aimage.png?table=block&amp;id=27cbbc89-34a4-8068-8d2f-d121dfda93fc&amp;t=27cbbc89-34a4-8068-8d2f-d121dfda93fc" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-text notion-block-27cbbc8934a480e18521c8827e189484">METR的研究分析了170项多样化的软件任务，涵盖从快速决策到复杂工程挑战等多个领域。该研究聚焦于三个核心维度：</div><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480589a24d61dd2d9dab1"><li>以人类表现作为基准进行对比</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480e59b07cc09b174d829"><li>衡量AI自主完成任务的能力</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480bea237d08c1d442cd1"><li>跟踪多个AI模型的性能演进趋势</li></ul><div class="notion-blank notion-block-27cbbc8934a4807ebfa1fa40b9f9caa2"> </div><div class="notion-text notion-block-27cbbc8934a480d29eeee43fda5e9a1b"><b>关键发现：AI能力快速增长</b></div><ul class="notion-list notion-list-disc notion-block-27cbbc8934a4802bb681c36b36b55234"><li><b>指数级扩展</b>：AI的能力正以远超传统计算进步的速度发展，使其能够执行过去需要人类认知与努力才能完成的任务。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480f19691d85e01b6e395"><li><b>持续加速</b>：自2019年以来，这种快速进步的模式持续被观察到，表明这是一个可靠的趋势，而非短暂的爆发。</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480c28db4dfcc257b339a"><li><b>性能里程碑</b>：当今的AI模型已能自主完成相当于人类近一小时工作量的任务——这与几年前相比是惊人的飞跃。</li></ul><div class="notion-blank notion-block-27cbbc8934a480ae95fadbdc4968b284"> </div><div class="notion-text notion-block-27cbbc8934a480728aafdafc135dd5d8"><b>AI目前擅长短时间任务</b></div><blockquote class="notion-quote notion-block-27cbbc8934a480a592c4d76fe7f6e9c0"><div>我们发现，人类专家完成任务所需的时间可以很好地预测模型在特定任务上的成功率：当前模型在人类耗时少于 4 分钟的任务中成功率几乎达到 100%，但在耗时超过 4 小时的任务中成功率不到 10%。这使我们能够通过“模型能够以 x% 的概率成功完成的任务长度（以人类为单位）”来表征给定模型的能力。</div></blockquote><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-27cbbc8934a480769211f435b348981d"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A10f5c30e-eae1-41dc-b849-96ddca936060%3Aimage.png?table=block&amp;id=27cbbc89-34a4-8076-9211-f435b348981d&amp;t=27cbbc89-34a4-8076-9211-f435b348981d" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-blank notion-block-27cbbc8934a480d994c5cee8cbcd86cc"> </div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-27cbbc8934a480a6aaa5d0010ec7ef44" data-id="27cbbc8934a480a6aaa5d0010ec7ef44"><span><div id="27cbbc8934a480a6aaa5d0010ec7ef44" class="notion-header-anchor"></div><a class="notion-hash-link" href="#27cbbc8934a480a6aaa5d0010ec7ef44" title="基于以上分析，AI相关开发者或企业，应重点关注："><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title"><b>基于以上分析，AI相关开发者或企业，应重点关注：</b></span></span></h3><ol start="1" class="notion-list notion-list-numbered notion-block-27cbbc8934a48036bf80fadbcb26b458" style="list-style-type:decimal"><li>聚焦当前AI优势</li></ol><div class="notion-text notion-block-27cbbc8934a480d48a9ed577410dab4a">与其为推测中的未来AI能力设计方案，不如充分利用现有优势。构建专注于1小时时限内任务的AI智能体，在确保即时商业影响的同时为未来扩展性做好准备。（任务越短，AI成功率会越高）</div><ol start="1" class="notion-list notion-list-numbered notion-block-27cbbc8934a480b0bce7cb6fa2b33739" style="list-style-type:decimal"><li>设计可扩展架构（基模能力会越来越强）</li></ol><div class="notion-text notion-block-27cbbc8934a480d98419ff556b02be7f">随着AI能力提升，模块化AI架构将至关重要。构建能随底层模型升级无缝扩展复杂度的AI智能体，使当今的30分钟AI解决方案无需重构即可进化成多小时、多智能体系统。</div><ol start="1" class="notion-list notion-list-numbered notion-block-27cbbc8934a480618abdee9ea4a4b104" style="list-style-type:decimal"><li>分解复杂工作流（化繁为简）</li></ol><div class="notion-text notion-block-27cbbc8934a48040a1f8f55e1de78aa8">鉴于AI目前擅长处理1小时内任务，可将更长流程拆分为可管理段落。这种结构化方法在提升效率的同时，为AI持续进化奠定基础。</div><ol start="1" class="notion-list notion-list-numbered notion-block-27cbbc8934a48054a625d30b0a387ba8" style="list-style-type:decimal"><li>准备多智能体协作</li></ol><div class="notion-text notion-block-27cbbc8934a4803f88cdeeaa9266029c">未来不仅需要更聪明的单个AI智能体——更需要协同合作的AI生态系统。多智能体系统将突破单任务限制，通过协同合作无缝执行更复杂的多阶段工作流。</div><div class="notion-blank notion-block-27cbbc8934a480fba1dcdc35a9bc24fc"> </div><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-27cbbc8934a48057aec2cb1067259453" data-id="27cbbc8934a48057aec2cb1067259453"><span><div id="27cbbc8934a48057aec2cb1067259453" class="notion-header-anchor"></div><a class="notion-hash-link" href="#27cbbc8934a48057aec2cb1067259453" title="参考："><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">参考：</span></span></h3><div class="notion-text notion-block-27cbbc8934a480dabc5cd23cd778df0b"><span class="notion-link-mention"><a href="https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/" target="_blank" rel="noopener noreferrer" class="notion-link-mention-link"><img class="notion-link-mention-icon" src="https://metr.org/assets/images/favicon/favicon.png"/><span class="notion-link-mention-title">Measuring AI Ability to Complete Long Tasks</span></a><div class="notion-link-mention-preview"><article class="notion-link-mention-card"><img class="notion-link-mention-preview-thumbnail" src="https://metr.org/assets/images/measuring-ai-ability-to-complete-long-tasks/length-of-tasks-log.png" alt="Measuring AI Ability to Complete Long Tasks" referrerPolicy="same-origin"/><div class="notion-link-mention-preview-content"><p class="notion-link-mention-preview-title">Measuring AI Ability to Complete Long Tasks</p><p class="notion-link-mention-preview-description">We propose measuring AI performance in terms of the *length* of tasks AI agents can complete. We show that this metric has been consistently exponentially increasing over the past 6 years, with a doubling time of around 7 months. Extrapolating this trend predicts that, in under a decade, we will see AI agents that can independently complete a large fraction of software tasks that currently take humans days or weeks.</p><div class="notion-link-mention-preview-footer"><img class="notion-link-mention-preview-icon" src="https://metr.org/assets/images/favicon/favicon.png" referrerPolicy="same-origin"/><span class="notion-link-mention-preview-provider"></span></div></div></article></div></span></div><div class="notion-text notion-block-27cbbc8934a4803fa2e3cbc43993e759"><span class="notion-link-mention"><a href="https://arxiv.org/abs/2503.14499" target="_blank" rel="noopener noreferrer" class="notion-link-mention-link"><img class="notion-link-mention-icon" src="https://arxiv.org/static/browse/0.3.4/images/icons/apple-touch-icon.png" alt="arXiv.org"/><span class="notion-link-mention-provider">arXiv.org</span><span class="notion-link-mention-title">Measuring AI Ability to Complete Long Tasks</span></a><div class="notion-link-mention-preview"><article class="notion-link-mention-card"><img class="notion-link-mention-preview-thumbnail" src="https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png" alt="Measuring AI Ability to Complete Long Tasks" referrerPolicy="same-origin"/><div class="notion-link-mention-preview-content"><p class="notion-link-mention-preview-title">Measuring AI Ability to Complete Long Tasks</p><p class="notion-link-mention-preview-description">Despite rapid progress on AI benchmarks, the real-world meaning of benchmark performance remains unclear. To quantify the capabilities of AI systems in terms of human capabilities, we propose a...</p><div class="notion-link-mention-preview-footer"><img class="notion-link-mention-preview-icon" src="https://arxiv.org/static/browse/0.3.4/images/icons/apple-touch-icon.png" alt="arXiv.org" referrerPolicy="same-origin"/><span class="notion-link-mention-preview-provider">arXiv.org</span></div></div></article></div></span></div><div class="notion-text notion-block-27cbbc8934a480c5a9fde4e874b3d752"><span class="notion-link-mention"><a href="https://www.linkedin.com/pulse/ai-capabilities-doubling-every-7-months-you-keeping-up-vijay-chintha-ma04e/" target="_blank" rel="noopener noreferrer" class="notion-link-mention-link"><img class="notion-link-mention-icon" src="https://static.licdn.com/aero-v1/sc/h/al2o9zrvru7aqj8e1x2rzsrca" alt="LinkedInEditors"/><span class="notion-link-mention-provider">LinkedInEditors</span><span class="notion-link-mention-title">AI Capabilities are Doubling Every 7 Months. Are You Keeping Up?</span></a><div class="notion-link-mention-preview"><article class="notion-link-mention-card"><img class="notion-link-mention-preview-thumbnail" src="https://media.licdn.com/dms/image/v2/D4E12AQFlCrgPGDhRcg/article-cover_image-shrink_720_1280/B4EZXbweuXHUAI-/0/1743148679298?e=2147483647&amp;v=beta&amp;t=AZTp7ci9tCWGH6q-YH9yLWK2iwZGa8oCoZQl2ezefRM" alt="AI Capabilities are Doubling Every 7 Months. Are You Keeping Up?" referrerPolicy="same-origin"/><div class="notion-link-mention-preview-content"><p class="notion-link-mention-preview-title">AI Capabilities are Doubling Every 7 Months. Are You Keeping Up?</p><p class="notion-link-mention-preview-description">In the rapidly evolving landscape of artificial intelligence (AI), we are witnessing an unprecedented acceleration in AI&#x27;s capabilities, fundamentally reshaping our understanding of technological progress. While Moore’s Law has historically guided expectations for computing advancements, recent rese</p><div class="notion-link-mention-preview-footer"><img class="notion-link-mention-preview-icon" src="https://static.licdn.com/aero-v1/sc/h/al2o9zrvru7aqj8e1x2rzsrca" alt="LinkedInEditors" referrerPolicy="same-origin"/><span class="notion-link-mention-preview-provider">LinkedInEditors</span></div></div></article></div></span></div></main></div>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[让每个故事都成为孩子的专属世界：用 Storybook 点燃个性化学习的火花]]></title>
            <link>iqiujin.com/article/24bbbc89-34a4-801b-ae4c-efdf23803616</link>
            <guid>iqiujin.com/article/24bbbc89-34a4-801b-ae4c-efdf23803616</guid>
            <pubDate>Sun, 10 Aug 2025 00:00:00 GMT</pubDate>
            <description><![CDATA[一定要将孩子的个性化带入到故事，一定要将想传达的内容带入到故事。]]></description>
            <content:encoded><![CDATA[<div id="notion-article" class="mx-auto overflow-hidden "><main class="notion light-mode notion-page notion-block-24bbbc8934a4801bae4cefdf23803616"><div class="notion-viewport"></div><div class="notion-collection-page-properties"></div><div class="notion-callout notion-gray_background_co notion-block-24bbbc8934a48060a8f9cc0ddac765a4"><div class="notion-page-icon-inline notion-page-icon-span"><span class="notion-page-icon" role="img" aria-label="💡">💡</span></div><div class="notion-callout-text"><div class="notion-text notion-block-24bbbc8934a4802ebd93da2285f274c3">一定要将孩子的个性化带入到故事，一定要将想传达的内容带入到故事。</div></div></div><hr class="notion-hr notion-block-24bbbc8934a48063ba45de229201b0c0"/><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-24bbbc8934a48013ba79f650f5ae347a" data-id="24bbbc8934a48013ba79f650f5ae347a"><span><div id="24bbbc8934a48013ba79f650f5ae347a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a48013ba79f650f5ae347a" title="✍️ 引言"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">✍️ 引言</span></span></h3><div class="notion-text notion-block-24bbbc8934a48043b982c2b4d9edc562">想让阅读不仅仅是听故事，而是成为孩子探索自我、认知世界的个性化旅程吗？答案就藏在“个性化”这把钥匙里。借助 Storybook 工具，我们可以将每一次阅读都打造成独一无二的互动学习体验，让孩子真正成为故事的主角。</div><hr class="notion-hr notion-block-24bbbc8934a480a08027dd1dc8248f55"/><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-24bbbc8934a4806c987beb7e3ed1340a" data-id="24bbbc8934a4806c987beb7e3ed1340a"><span><div id="24bbbc8934a4806c987beb7e3ed1340a" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a4806c987beb7e3ed1340a" title="🎯 核心要点"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">🎯 核心要点</span></span></h3><div class="notion-text notion-block-24bbbc8934a480cca164c2ed7d1f2d1e">本文将从<b>互动提问、现实连接、创意延伸</b>三个维度，深度解析如何运用 Storybook，将孩子的<b>个性化特质</b>与<b>家长的教育目标</b>巧妙融合，让每一次阅读都成为一场专属的学习盛宴。</div><hr class="notion-hr notion-block-24bbbc8934a4802eb5b0c1a2a6ea91a3"/><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-24bbbc8934a480a7bedaebbc52ba74d3" data-id="24bbbc8934a480a7bedaebbc52ba74d3"><span><div id="24bbbc8934a480a7bedaebbc52ba74d3" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a480a7bedaebbc52ba74d3" title="📖 文章正文"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">📖 文章正文</span></span></h3><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-24bbbc8934a48055ad0ce116a54e0f7e" data-id="24bbbc8934a48055ad0ce116a54e0f7e"><span><div id="24bbbc8934a48055ad0ce116a54e0f7e" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a48055ad0ce116a54e0f7e" title="1. 打造专属剧本：互动与提问的个性化升级"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">1. 打造专属剧本：互动与提问的个性化升级</span></span></h4><div class="notion-text notion-block-24bbbc8934a48015a5a7edd267f0e1c5">在 Storybook 的世界里，主角就是孩子自己，因此每一个提问都应是与孩子内心的直接对话。</div><ul class="notion-list notion-list-disc notion-block-24bbbc8934a48081965debc57fbae0fb"><li>注入“你”的元素</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a48081965debc57fbae0fb"><div class="notion-text notion-block-24bbbc8934a480c08b87f7da54eee2b4">告别千篇一律的提问。在阅读前，直接将孩子融入故事：“看，这个小探险家和你一样也叫 [孩子的名字]！你猜，他会在这个神奇的森林里发现什么？”或者“这个小精灵的情绪看起来有点低落，你觉得发生什么事了？如果是 [孩子的名字]，你会怎么安慰他？”</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480548088f1ea716268ae"><li>定制化预测与决策</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480548088f1ea716268ae"><div class="notion-text notion-block-24bbbc8934a48022bc0aeedb6e9077bb">针对孩子的性格和兴趣点提问。如果孩子热爱画画，可以问：“如果你是这个故事的插画师，你会把下一页画成什么样子？”如果孩子逻辑性强，可以问：“你觉得 [孩子的名字] 这个决定聪明吗？有没有更好的办法？”</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480158670c4bdcb91231e"><li>“我的”故事复述</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480158670c4bdcb91231e"><div class="notion-text notion-block-24bbbc8934a4806c8110c44dc633b0de">鼓励孩子用第一人称“我”来复述故事。这不仅仅是记忆力训练，更是引导他们将自己代入角色，体验角色的情感与成长，从而深刻理解故事的核心。</div></ul></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-24bbbc8934a48035a482e8a39352e007" data-id="24bbbc8934a48035a482e8a39352e007"><span><div id="24bbbc8934a48035a482e8a39352e007" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a48035a482e8a39352e007" title="2. 连接独一无二的现实：让故事在孩子的生活中扎根"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">2. 连接独一无二的现实：让故事在孩子的生活中扎根</span></span></h4><div class="notion-text notion-block-24bbbc8934a480adb380d204211c539c">故事的真正力量，在于它能成为孩子理解自己、解决现实问题的“工具箱”。</div><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4809abb5ae4fb4f9f1b90"><li>锚定个人经历</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4809abb5ae4fb4f9f1b90"><div class="notion-text notion-block-24bbbc8934a480c1aa9fd4abf978d0d1">将故事中的情节与孩子的亲身经历紧密相连。读到分享玩具的情节时，可以回忆并赞扬孩子上次与小伙伴分享的经历：“记得你上次把最喜欢的积木分给小明玩吗？你和书里的小英雄 [孩子的名字] 一样慷慨！”这种积极的联结，能极大地增强孩子的自我认同感。</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4806cbd71c892f2affd97"><li>解决“我的”难题</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4806cbd71c892f2affd97"><div class="notion-text notion-block-24bbbc8934a4800bae0cd26fc1414bfc">当孩子面临特定的情绪或挑战时（例如，害怕去幼儿园、与朋友争吵），为他量身寻找或创作一个故事。通过故事里**[孩子的名字]**的经历，孩子可以“预演”解决方案，学习如何面对和处理自己的困惑。现在，借助 AI 故事生成工具，你甚至可以输入孩子的名字、困扰和兴趣点，生成一个完全属于他的、具有疗愈功能的小故事。</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4801b98d5e19ff499ae48"><li>引入“我感兴趣”的新知</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4801b98d5e19ff499ae48"><div class="notion-text notion-block-24bbbc8934a4802d8048cfa1947c1676">当故事中出现孩子着迷的概念时（比如恐龙、太空、魔法），顺势展开一场个性化的深度探索。可以一起制作一个关于“我的星球”的黏土模型，或者画一本属于自己的“恐龙百科全书”，让知识的学习过程充满个人创造的烙印。</div></ul></ul><h4 class="notion-h notion-h3 notion-h-indent-1 notion-block-24bbbc8934a480678648efca81d795e3" data-id="24bbbc8934a480678648efca81d795e3"><span><div id="24bbbc8934a480678648efca81d795e3" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a480678648efca81d795e3" title="3. 延伸至无限的创造：从故事到个性化游戏"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">3. 延伸至无限的创造：从故事到个性化游戏</span></span></h4><div class="notion-text notion-block-24bbbc8934a480fcb0aed273e0fa7751">将故事的边界从书本延伸到无限的现实世界，让孩子的想象力和创造力得到尽情释放。</div><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4805ba3b2f2a79402bc3a"><li>定制化角色扮演</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a4805ba3b2f2a79402bc3a"><div class="notion-text notion-block-24bbbc8934a48061ab30e69599ca33f9">不只是扮演已有角色，更可以创造新角色！“如果故事里有你的加入，你会是谁？你想拥有什么超能力？你想和主角 [孩子的名字] 一起去哪里探险？”这能激发孩子更高层次的创造性思维。</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a48035b5dbd69de662905e"><li>“我的”手工坊</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a48035b5dbd69de662905e"><div class="notion-text notion-block-24bbbc8934a480859a48cb06f1987805">根据故事内容，结合孩子的动手能力和兴趣，设计独一无二的手工活动。读完一个关于海洋的故事后，可以让他用捡来的贝壳和石头，搭建一个属于自己的“海底城堡”。</div></ul></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480d9a049ff4665c82f0c"><li>成为小小创作者</li><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480d9a049ff4665c82f0c"><div class="notion-text notion-block-24bbbc8934a480618fa4c58ce420544b">终极的个性化，是鼓励孩子创作自己的故事书。让他们自由地绘画、叙述，由您来记录。这不仅是创造力的锻炼，更是对其个人想法和声音的尊重与肯定，是送给孩子最宝贵的自信心礼物。</div></ul></ul><hr class="notion-hr notion-block-24bbbc8934a4803aa18cc24b8a7ebdc5"/><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-24bbbc8934a4801680edd973d42bb8d6" data-id="24bbbc8934a4801680edd973d42bb8d6"><span><div id="24bbbc8934a4801680edd973d42bb8d6" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a4801680edd973d42bb8d6" title="💡 关键洞察"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">💡 关键洞察</span></span></h3><ol start="1" class="notion-list notion-list-numbered notion-block-24bbbc8934a4802ca01be63284e48b0d" style="list-style-type:decimal"><li><b>个性化是打开心门的钥匙</b>：只有将故事与孩子自身的兴趣、特质相结合，他们才会真正投入。</li></ol><ol start="2" class="notion-list notion-list-numbered notion-block-24bbbc8934a4808baf6ef11b191194dc" style="list-style-type:decimal"><li><b>教育意图的“隐形”传递</b>：最有效的教育不是说教，而是将教育目标巧妙地融入故事中，让孩子自己去发现和领悟。</li></ol><ol start="3" class="notion-list notion-list-numbered notion-block-24bbbc8934a480aeab82dd5060e5c689" style="list-style-type:decimal"><li><b>“做”故事比“听”故事更有效</b>：通过延伸活动，将故事从书本带入现实，让学习成为一种有趣的实践。</li></ol><hr class="notion-hr notion-block-24bbbc8934a4806b8949f1215b9aee06"/><h3 class="notion-h notion-h2 notion-h-indent-0 notion-block-24bbbc8934a480f5ae5fd94caeb762cf" data-id="24bbbc8934a480f5ae5fd94caeb762cf"><span><div id="24bbbc8934a480f5ae5fd94caeb762cf" class="notion-header-anchor"></div><a class="notion-hash-link" href="#24bbbc8934a480f5ae5fd94caeb762cf" title="🚀 行动建议"><svg viewBox="0 0 16 16" width="16" height="16"><path fill-rule="evenodd" d="M7.775 3.275a.75.75 0 001.06 1.06l1.25-1.25a2 2 0 112.83 2.83l-2.5 2.5a2 2 0 01-2.83 0 .75.75 0 00-1.06 1.06 3.5 3.5 0 004.95 0l2.5-2.5a3.5 3.5 0 00-4.95-4.95l-1.25 1.25zm-4.69 9.64a2 2 0 010-2.83l2.5-2.5a2 2 0 012.83 0 .75.75 0 001.06-1.06 3.5 3.5 0 00-4.95 0l-2.5 2.5a3.5 3.5 0 004.95 4.95l1.25-1.25a.75.75 0 00-1.06-1.06l-1.25 1.25a2 2 0 01-2.83 0z"></path></svg></a><span class="notion-h-title">🚀 行动建议</span></span></h3><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480fe8de7d7ee2c5a1448"><li><b>提问升级</b>：下次给孩子讲故事时，把“你猜猜”升级为“<b>你猜猜</b> <b>[孩子的名字]</b> <b>会怎么样</b>”。</li></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480468990c8704b071bbe"><li><b>目标植入</b>：在选择故事时，思考你希望通过它传递什么具体的价值观或知识。</li></ul><ul class="notion-list notion-list-disc notion-block-24bbbc8934a480baa7c9d2a69c6764a4"><li><b>实践联动</b>：读完故事后，和孩子一起根据故事内容做一个手工或玩一个游戏。</li></ul><div class="notion-blank notion-block-24bbbc8934a4802ea9d8fad3ab2174ce"> </div><figure class="notion-asset-wrapper notion-asset-wrapper-image notion-block-24bbbc8934a48004bc52dc1813b8e573"><div style="position:relative;display:flex;justify-content:center;align-self:center;width:100%;max-width:100%;flex-direction:column;height:100%"><img style="object-fit:cover" src="https://www.notion.so/image/attachment%3A65b219a6-1c11-4137-b93b-c2cadcb5f8c2%3Aimage.png?table=block&amp;id=24bbbc89-34a4-8004-bc52-dc1813b8e573&amp;t=24bbbc89-34a4-8004-bc52-dc1813b8e573" alt="notion image" loading="lazy" decoding="async"/></div></figure><div class="notion-blank notion-block-27cbbc8934a480c98413c9e5074d52ab"> </div><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480b3858bd73e55712f69"><li><b>Help your child understand a complex topic:</b> Create a story that explains the solar system to my 5 year old.</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a48021b523f97d9dd582df"><li><b>Teach a lesson through storytelling:</b> Teach a 7-year-old boy about the importance of being kind to his little brother. My son loves elephants so let’s make the main character an elephant.</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a48072a812d833efeba418"><li><b>Bring personal artwork to life</b>: Upload an image of a kid&#x27;s drawing and modify this example prompt for your use case: &quot;This is my kid’s drawing. He’s 7 years old. Write a creative storybook that brings his drawing to life.”</li></ul><ul class="notion-list notion-list-disc notion-block-27cbbc8934a480b88c80cd7836ed15d9"><li><b>Turn memories into magical stories:</b> Upload photos from your family trip to Paris and create a personalized adventure.</li></ul></main></div>]]></content:encoded>
        </item>
    </channel>
</rss>