7wu1wm0 发表于 2024-9-27 16:35:58

写给小白的AI入门科普


    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> 到底什么是AI?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI,是artificial intelligence的缩写。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Artificial,<span style="color: black;">非常多</span><span style="color: black;">朋友</span>认字认半边,会以为是艺术(art)的什么形容词。<span style="color: black;">并不</span>然,artificial的意思<span style="color: black;">便是</span>“人工的、人造的”,和natural(天然的)是反义词。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Intelligence,这个<span style="color: black;">不易</span>认错,是“智能”的意思。英特尔(Intel)<span style="color: black;">机构</span>的名字,<span style="color: black;">便是</span>基于这个词的前五个字母。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">结合起来,AI,<span style="color: black;">便是</span>“人工的、人造的智能”,用人为的手段,创造智能。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-axegupay5k/152476af9c754f1cb4368ad32ee34a0b~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=3QNZVshW55mDzkpHQj7VspEQSJQ%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">关于AI的定义,行业里有<span style="color: black;">非常多</span>种说法。比较学术化的一种,是这么说的:</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">AI,是<span style="color: black;">科研</span>、<span style="color: black;">研发</span>用于模拟、延伸和扩展人的智能<span style="color: black;">行径</span>的理论、<span style="color: black;">办法</span>、技术及应用系统的一门综合性科学。</span></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这个定义很拗口,看得都头大。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">其实,</span><span style="color: black;">针对</span>AI,<span style="color: black;">咱们</span><span style="color: black;">能够</span>拆解来看。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">首要</span>,AI的本质属性,是一门<strong style="color: blue;">科学</strong>,是一个技术<span style="color: black;">行业</span>。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">它<span style="color: black;">触及</span>到了计算机科学、数学、统计学、哲学、心理学等多种学科的知识,但总体上,归类于<strong style="color: blue;">计算机学科</strong>之下。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">其次,AI的<span style="color: black;">科研</span>目的,是让一个“系统<span style="color: black;">”</span>具备智能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这个<span style="color: black;">“系统”</span>,<span style="color: black;">能够</span>是一套软件程序,<span style="color: black;">亦</span>可以是一台计算机,<span style="color: black;">乃至</span>是一个<span style="color: black;">设备</span>人。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">第三,什么样的水平,才叫做真正的智能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这是问题的关键。<span style="color: black;">日前</span>看来,能够像人<span style="color: black;">同样</span>感知、理解、思考、判断、决策,<span style="color: black;">便是</span>实现了人工智能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">协同</span><span style="color: black;">设备</span>人、机械臂等<span style="color: black;">理学</span>载体,AI<span style="color: black;">亦</span><span style="color: black;">能够</span>实现行动能力。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">综合以上三点,理解AI的定义就比较容易了。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> AI和普通计算机有什么区别?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI,<span style="color: black;">日前</span>仍然是基于计算机的基本玩法,采用的是半导体芯片技术(<span style="color: black;">因此</span>经常会被<span style="color: black;">叫作</span>为“硅基”),以及计算机的<span style="color: black;">有些</span>体系和平台。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">那样</span>,它和传统的计算机程序,有什么区别呢?</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">传统的计算机程序,<span style="color: black;">便是</span>一个规则的集合。程序员<span style="color: black;">经过</span>代码<span style="color: black;">通知</span>计算机规则,计算机<span style="color: black;">按照</span>规则,对输入数据进行判断和处理。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">例如经典的“if……else……(<span style="color: black;">倘若</span>……否则……)”语句——“<span style="color: black;">倘若</span>大于65岁,就<span style="color: black;">离休</span>。否则,继续上班”。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">而后</span>,计算机程序会<span style="color: black;">按照</span>这个规则,对所有输入年龄数据进行判断和处理。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">然则</span>,在现实生活中,<span style="color: black;">非常多</span>要素(例如图像和声音)是极为<span style="color: black;">繁杂</span>和多样的,<span style="color: black;">咱们</span>很难给出固定的规则,让计算机实现高准确率的判断和处理。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">例如,判断一只狗是不是狗。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">狗有<span style="color: black;">非常多</span>品种,每种狗有<span style="color: black;">区别</span>的颜色、体型、五官特征。狗在<span style="color: black;">区别</span>的时间,<span style="color: black;">亦</span>有<span style="color: black;">区别</span>的表情、姿势。狗还会<span style="color: black;">处在</span><span style="color: black;">区别</span>的背景环境下。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/e8e12dd401294d61beae6d6a3cefc4e5~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=yRtyI2CsafdPtyHz3xVkCx3FhqI%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">因此</span>,计算机<span style="color: black;">经过</span>摄像头<span style="color: black;">捉捕</span>到的狗的影像,是无穷尽的。很难<span style="color: black;">经过</span>有限数量的规则,去<span style="color: black;">帮忙</span>计算机做出判断。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">想要让计算机实现像人<span style="color: black;">同样</span>的智能,<span style="color: black;">不可</span>采用简单的规则驱动,而是应该像教孩童<span style="color: black;">同样</span>,<span style="color: black;">持续</span>输入数据和答案,让他<span style="color: black;">自动</span>总结特征,形成自己的判断规则。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/2ddc61b0ca4b41b2b7f518d92097f8b5~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=Y%2BSQ%2BCI8EV7POCHG%2BEmBIatE0zY%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">换言之,在经典的程序设计中,人们输入的是规则(即程序)和数据,系统输出的是答案。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/e546432b43ba43f7a345daf275797576~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=YoTZDtEE%2BoA9X9v6Fe7Qqc3L4c0%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">而AI的计算过程,分为两个<span style="color: black;">过程</span>:</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">第1</span>个<span style="color: black;">过程</span>,输入的是数据和预期得到的答案,系统输出的是规则。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">第二个<span style="color: black;">过程</span>,将输出的规则应用于新的数据,<span style="color: black;">而后</span>再输出答案。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p26-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/6c9527a98a744df98a7c5d97a6ddf231~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=LpqSEMdHUeTalHZXsapp%2FSFDhOs%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">第1</span>步,<span style="color: black;">咱们</span><span style="color: black;">能够</span><span style="color: black;">叫作</span>之为“训练”。第二步,才是真正“干活”。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">这<span style="color: black;">便是</span>传统计算程序和<span style="color: black;">此刻</span>主流AI技术的一个典型区别。(<span style="color: black;">重视</span>,我说的是“<span style="color: black;">此刻</span>主流AI”。有<span style="color: black;">有些</span>“历史AI”和“非主流AI”,玩法不<span style="color: black;">同样</span>。<span style="color: black;">不可</span>一概而论。)</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> AI,有<span style="color: black;">那些</span>类别?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">前面说了,人工智能是一个非常庞大的科学<span style="color: black;">行业</span>。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">从1950年代正式诞生<span style="color: black;">败兴</span>,围绕人工智能,<span style="color: black;">已然</span>有<span style="color: black;">非常多</span><span style="color: black;">专家</span>进行了<span style="color: black;">海量</span>的研究,<span style="color: black;">亦</span>输出了<span style="color: black;">非常多</span>非常了不起的成果。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这些<span style="color: black;">科研</span>,<span style="color: black;">按照</span>思路方向的<span style="color: black;">区别</span>,被分为了<span style="color: black;">非常多</span>种学派。比较有<span style="color: black;">表率</span>性的,是符号主义学派、联结主义学派、<span style="color: black;">行径</span>主义学派。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这些学派并<span style="color: black;">无</span>对错之分,相互之间<span style="color: black;">亦</span>有<span style="color: black;">有些</span>交叉融合。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">初期</span>的时候(1960-1990),符号主义(以专家系统、知识图谱为<span style="color: black;">表率</span>)是主流。后来,从1980年<span style="color: black;">起始</span>,联结主义(以神经网络为<span style="color: black;">表率</span>)崛起,<span style="color: black;">始终</span>到<span style="color: black;">此刻</span>,都是主流。</p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p6-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/8dad8653208c47bca89526215df052ef~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=a0ur%2B2wsxqPgILw16ttFZYfi3jc%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">将来,<span style="color: black;">亦</span>许有新的技术崛起,形成新的学派,<span style="color: black;">亦</span>不<span style="color: black;">必定</span>。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">除了方向路线之外,<span style="color: black;">咱们</span><span style="color: black;">亦</span><span style="color: black;">能够</span>从智能水平以及应用<span style="color: black;">行业</span>等方面对AI进行<span style="color: black;">归类</span>。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">按智能水平,<span style="color: black;">能够</span>分为:<strong style="color: blue;">弱人工智能(Weak AI</strong><span style="color: black;"><strong style="color: blue;">)</strong></span>、<strong style="color: blue;">强人工智能</strong><span style="color: black;"><strong style="color: blue;">(</strong></span><strong style="color: blue;">Strong AI</strong><span style="color: black;"><strong style="color: blue;">)</strong></span>、<span style="color: black;"><strong style="color: blue;">超人工智能(Super AI)</strong>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">弱人工智能只专精于单一任务或一组<span style="color: black;">关联</span>的任务,不具备通用智能能力。<span style="color: black;">咱们</span><span style="color: black;">日前</span>就<span style="color: black;">处在</span>这个<span style="color: black;">周期</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">强人工智能更厉害<span style="color: black;">有些</span>,<span style="color: black;">拥有</span><span style="color: black;">必定</span>的通用智能能力,能够理解、学习并应用于<span style="color: black;">各样</span><span style="color: black;">区别</span>的任务。这个还<span style="color: black;">处在</span>理论和<span style="color: black;">科研</span><span style="color: black;">周期</span>,还没落地。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">超人工智能当然是最强的。它在几乎所有方面都超过人类智能,<span style="color: black;">包含</span>创造力、社交技能等。超人工智能是<span style="color: black;">将来</span>的终极形态,<span style="color: black;">咱们</span>假设它能够实现。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">关于按应用<span style="color: black;">行业</span>的AI<span style="color: black;">归类</span>,<span style="color: black;">咱们</span>待会再说。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> 什么是<span style="color: black;">设备</span>学习?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">其实<span style="color: black;">咱们</span>前面介绍规则总结的时候,其实<span style="color: black;">已然</span><span style="color: black;">说到</span>了<span style="color: black;">设备</span>学习。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">设备</span>学习的核心思想,是构建一个<span style="color: black;">能够</span>从数据中学习的模型,并利用这个模型来进行预测或决策。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">设备</span>学习不是一个<span style="color: black;">详细</span>的模型或算法。它<span style="color: black;">包含</span>了<span style="color: black;">非常多</span>种类型,例如:</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;">监督学习</strong>:算法从带有标签的数据集中学习,即<span style="color: black;">每一个</span>训练样本都有一个已知的结果。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;">无监督学习</strong>:算法从<span style="color: black;">无</span>标签的数据集中学习。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;">半监督学习</strong>:结合了少量的带标签数据和<span style="color: black;">海量</span>的未带标签数据进行训练。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;">强化学习</strong>:<span style="color: black;">经过</span>试错的方式,学习<span style="color: black;">那些</span><span style="color: black;">行径</span><span style="color: black;">能够</span><span style="color: black;">得到</span>奖励,<span style="color: black;">那些</span><span style="color: black;">行径</span>会<span style="color: black;">引起</span><span style="color: black;">处罚</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> </strong></span><strong style="color: blue;">什么是深度学习?</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">深度学习,<span style="color: black;">详细</span><span style="color: black;">来讲</span>,是深度神经网络学习。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">深度学习是<span style="color: black;">设备</span>学习的一个重要分支。<span style="color: black;">设备</span>学习底下有一条“神经网络”路线,而深度学习,是加强版的“神经网络”学习。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">神经网络是联结主义的<span style="color: black;">表率</span>。顾名思义,这个路线是模仿人脑的工作原理,<span style="color: black;">创立</span>神经元之间的联结模型,以此实现人工神经运算。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/2aa22404a9b748739785a33a54ddaf89~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=IaBfaFNtzQSkTyZJ%2Fm%2FeNnM3SSY%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">深度学习<span style="color: black;">所说</span>的“深度”,是神经网络中<strong style="color: blue;">“<span style="color: black;">隐匿</span>层”</strong>的层级。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">经典<span style="color: black;">设备</span>学习算法<span style="color: black;">运用</span>的神经网络,<span style="color: black;">拥有</span>输入层、一个或两个“<span style="color: black;">隐匿</span>层”和一个输出层。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">深度学习算法<span style="color: black;">运用</span>了<span style="color: black;">更加多</span>的“<span style="color: black;">隐匿</span>层”(数百个)。它的能力更加强大,让神经网络能够完成更困难的工作。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">设备</span>学习、神经网络和深度学习的关系,<span style="color: black;">经过</span>下面的图<span style="color: black;">能够</span>看出:</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/b2b9c44869d748f08ded3a63ee0c7eaa~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=cYTmDrC53lZXHDPg4TVeLgwUINI%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> 什么是卷积神经网络、循环神经网络?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">神经网络从1980年代<span style="color: black;">起始</span>崛起之后,就形<span style="color: black;">成为了</span><span style="color: black;">非常多</span>的模型和算法。<span style="color: black;">区别</span>的模型和算法,有着各自的特性和功能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">卷积神经网络(Convolutional Neural Network,CNN)和循环神经网络(Recurrent Neural Network,RNN),是1990年代<span style="color: black;">上下</span>诞生的比较知名的神经网络模型。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">它们的<span style="color: black;">详细</span>工作原理比较<span style="color: black;">繁杂</span>。反正<span style="color: black;">大众</span>记住:</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">卷积神经网络(CNN)是一种用于处理<span style="color: black;">拥有</span>类似网格结构的数据(例如图像和视频)的神经网络。<span style="color: black;">因此</span>,它<span style="color: black;">一般</span>用于计算机视觉中,<span style="color: black;">能够</span>用来<strong style="color: blue;">图像识别和图像<span style="color: black;">归类</span></strong>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">而循环神经网络(RNN)是一种用于处理序列数据的神经网络,例如语言模型和时间序列预测。<span style="color: black;">因此</span>,它<span style="color: black;">一般</span>用于<strong style="color: blue;">自然语言处理和语音识别</strong>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> 什么是transformer?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">transformer<span style="color: black;">亦</span>是一个神经网络模型。它比卷积神经网络和循环神经网络更加<span style="color: black;">青年</span>(2017年由谷歌<span style="color: black;">科研</span>团队提出),<span style="color: black;">亦</span>更加强大。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">做为</span>非专业人士,不需要去<span style="color: black;">科研</span>它的工作原理,只需要<span style="color: black;">晓得</span>:</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">1、它是一种深度学习模型;</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">2、它<span style="color: black;">运用</span>了一种名为<strong style="color: blue;">自<span style="color: black;">重视</span>力(self-attention)</strong>的机制;</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">3、它有效<span style="color: black;">处理</span>了卷积神经网络和循环神经网络的瓶颈(局限性)问题;</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">4、它很适合自然语言处理(NLP)任务。相比循环神经网络,它的计算<span style="color: black;">能够</span>高度并行化,简化了模型架构,训练效率<span style="color: black;">亦</span>大大<span style="color: black;">提高</span>;</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">5、它<span style="color: black;">亦</span>被扩展到了其他<span style="color: black;">行业</span>,如计算机视觉和语音识别。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">6、<span style="color: black;">此刻</span><span style="color: black;">咱们</span>经常<span style="color: black;">说到</span>的大模型,几乎都是以<span style="color: black;">transformer为<span style="color: black;">基本</span>。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/fa6ed0cf37e04ff3a9c4e0cacb60c9e6~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=wXscxVeRxqKlCeJ%2FkFrU3qnzL8s%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">神经网络还有<span style="color: black;">非常多</span>种,我在网上找到一张图,供参考:</p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/6e184cdaad0740128ecfe7eddb0fa9c5~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=g0UhZve%2BG%2BD4jRbPquU1bW%2Bbh%2FM%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> </strong></span><strong style="color: blue;">什么是大模型?</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这两年AI火,<span style="color: black;">便是</span><span style="color: black;">由于</span>大模型火。<span style="color: black;">那样</span>,什么是大模型?</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">大模型,<span style="color: black;">是<span style="color: black;">拥有</span>庞大<strong style="color: blue;">参数</strong>规模和<span style="color: black;">繁杂</span>计算结构的<strong style="color: blue;"><span style="color: black;">设备</span>学习模型</strong>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">参数,<span style="color: black;">指的是</span>在模型训练过程中,学习和<span style="color: black;">调节</span>的变量。参数定义了模型的<span style="color: black;">行径</span>、性能、实现的成本以及对计算资源的<span style="color: black;">需要</span>。简单<span style="color: black;">来讲</span>,参数是模型内部用来做出预测或决策的部分。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">大模型,<span style="color: black;">一般</span><span style="color: black;">持有</span>数百万至数十亿的参数。相对应的,参数少的,<span style="color: black;">便是</span>小模型。对<span style="color: black;">有些</span>细分的<span style="color: black;">行业</span>或场景,小模型<span style="color: black;">亦</span>够用。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">大模型需要依赖大规模数据进行训练,对算力资源的消耗<span style="color: black;">极重</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">大模型有<span style="color: black;">非常多</span>种类别。<strong style="color: blue;"><span style="color: black;">一般</span>所说的大模型,<span style="color: black;">重点</span><span style="color: black;">指的是</span>语言大模型(以文本数据进行训练)。</strong>但<span style="color: black;">实质</span>上,还有视觉大模型(以图像数据进行训练),以及多模态大模型(文本和图像都有)。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">绝大<span style="color: black;">都数</span>大模型的<span style="color: black;">基本</span>核心结构,都是Transformer及其变体。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/f36783a06a3b4c878f5d18f226acd8d2~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=ePPaVs0btRXiPo8lHaZKo88IVUY%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">按应用<span style="color: black;">行业</span>,大模型<span style="color: black;">能够</span>分为通用大模型和行业大模型。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">通用大模型的训练数据集更加广泛,覆盖的<span style="color: black;">行业</span>更加全面。行业大模型,顾名思义,训练数据来自特定行业,应用于专门的<span style="color: black;">行业</span>(例如金融、医疗、法律、工业)。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span>GPT的本质是什么?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">GPT-1、GPT-2……GPT-4o,等等,都是美国OpenAI这家<span style="color: black;">机构</span>推出的语言大模型,<span style="color: black;">一样</span>都是基于Transformer架构。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">GPT的全<span style="color: black;">叫作</span>,叫做Generative Pre.trained Transformer,生成式-预训练-Transformer。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Generative(生成式),<span style="color: black;">暗示</span>该模型能够生成连续的、有<span style="color: black;">规律</span>的文本内容,<span style="color: black;">例如</span>完成对话、创作故事、编写代码<span style="color: black;">或</span>写诗写歌等。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">这儿</span>刚好提一下,<span style="color: black;">此刻</span>常说的AIGC,<span style="color: black;">便是</span>AI Generated Content,人工智能生成内容。内容,<span style="color: black;">能够</span>是文本、图像、音频、视频等。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">GPT系列面向文本,谷歌<span style="color: black;">亦</span>推出过竞品BERT。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">文生图,比较有代表性的是DALL·E(<span style="color: black;">亦</span>来自OpenAI)、Midjourney(知名度大)和Stable Diffusion(开源)。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">文生音频(音乐),有Suno(OpenAI)、Stable Audio Open(由Stability.ai开源)、Audiobox(Meta)。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">文生视频,有Sora(OpenAI)、Stable Video Diffusion(由Stability.ai开源)、Soya(开源)。图<span style="color: black;">亦</span><span style="color: black;">能够</span>生视频,例如腾讯的Follow-Your-Click。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/cc983cb490a946d4826ba8d2061addef~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=GNahUkXMm8Rc%2BcFLudaBFo2Fm2w%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">AIGC是一个“应用维度”的定义,它不是一个<span style="color: black;">详细</span>的技术或模型。AIGC的<span style="color: black;">显现</span>,扩展了AI的功能,<strong style="color: blue;">打破了此前AI<span style="color: black;">重点</span>用于识别的功能限制</strong>,拓宽了应用场景。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">好了,继续解释GPT的第二个字母——Pre.trained。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">Pre.trained(预训练),<span style="color: black;">暗示</span>该模型会先在一个大规模未标注文本语料库上进行训练,学习语言的统计规律和潜在结构。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">经过</span>预训练,模型才有了<span style="color: black;">必定</span>的通用性。训练的数据越庞大(如网页文本、<span style="color: black;">资讯</span>等),模型的能力就越强。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">大众</span><span style="color: black;">针对</span>AI的关注热潮,<span style="color: black;">重点</span>源于2023年初的ChatGPT爆火。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">ChatGPT的chat,是聊天的意思。ChatGPT是OpenAI基于GPT模型<span style="color: black;">研发</span>的一个AI对话应用服务(<span style="color: black;">亦</span><span style="color: black;">能够</span>理解为GPT-3.5)。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/07ee92b18a3c4617afa130b8511a98ee~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=vFd0roXRfB9ux8fbWJUCoz9DpRE%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">经过</span>这个服务,人们才<span style="color: black;">能够</span>亲身体验到GPT模型的强大,有利于技术的宣传和推广。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">事实证明,OpenAI的策略成功了。ChatGPT充分吸引了公众关注度,<span style="color: black;">亦</span>成功推动了AI<span style="color: black;">行业</span>的发展热潮。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> </strong></span><strong style="color: blue;">AI,<span style="color: black;">到底</span>能做什么?</strong></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI的<span style="color: black;">功效</span>,极为广泛。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">概括<span style="color: black;">来讲</span>,AI和传统计算机系统相比,能<span style="color: black;">供给</span>的拓展能力,<span style="color: black;">包含</span>:<strong style="color: blue;">图像识别</strong>、<strong style="color: blue;">语音识别</strong>、<strong style="color: blue;">自然语言处理</strong>、<strong style="color: blue;">具身智能</strong>等方面。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">图像识别,有时候<span style="color: black;">亦</span>被归类为计算机视觉(<span style="color: black;">Computer Vision,CV</span>),<span style="color: black;">让计算机具备理解和处理图像和视频的能力。<span style="color: black;">平常</span>的是摄像头、工业质检、人脸识别之类的。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/0de10ae4de3e4229baa9e71511eceaf7~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=lzq5WUCLjPxPqUoxl69Kv3KbMoQ%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">语音识别,<span style="color: black;">便是</span>理解和处理音频,<span style="color: black;">得到</span>音频所搭载的信息。<span style="color: black;">平常</span>的是手机语音助手、<span style="color: black;">tel</span>呼叫中心、声控智能家居之类的,多用于交互场景。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">自然语言处理,前面介绍过,<span style="color: black;">便是</span>使计算机能够理解和处理自然语言,<span style="color: black;">晓得</span><span style="color: black;">咱们</span>到底在说什么。这个很火,多用于创造性的工作,例如写<span style="color: black;">资讯</span>稿、写书面材料、视频制作、游戏<span style="color: black;">研发</span>、音乐创作等。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">具身智能,<span style="color: black;">便是</span>把人工智能搭载在一个<span style="color: black;">理学</span>形态(“<span style="color: black;">身体</span>”)上,<span style="color: black;">经过</span>与环境互动,来<span style="color: black;">得到</span>和展示智能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">带AI的<span style="color: black;">设备</span>人,属于具身智能。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">斯坦福大学年初推出的“Mobile ALOHA”,<span style="color: black;">便是</span>一个典型的家用具身<span style="color: black;">设备</span>人。它<span style="color: black;">能够</span>炒菜、煮咖啡<span style="color: black;">乃至</span>逗猫,<span style="color: black;">火热</span>全网。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">值得一提的是,并不是所有的<span style="color: black;">设备</span>人,都是人形<span style="color: black;">设备</span>人。<span style="color: black;">亦</span>不是所有的<span style="color: black;">设备</span>人,都用到了AI。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/e48ac51cce0f4627a48f96dffdc56db6~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=iGwUxLi5Nt2KvMm%2Bo5GZQ%2FftVuE%3D" style="width: 50%; margin-bottom: 20px;">
      <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">人形<span style="color: black;">设备</span>人</p>
    </div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">AI<span style="color: black;">尤其</span><span style="color: black;">善于</span>对海量数据进行处理,一方面<span style="color: black;">经过</span>海量数据进行学习和训练,另一方面,基于新的海量数据,完成人工<span style="color: black;">没法</span>完成的工作。<span style="color: black;">或</span>说,找到海量数据中潜在的规律。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">日前</span>AI在社会各个垂直行业的应用,<span style="color: black;">重点</span>是围绕上面的能力进行延展。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span>举<span style="color: black;">有些</span><span style="color: black;">平常</span>的例子。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在医疗<span style="color: black;">行业</span>,<span style="color: black;">AI<span style="color: black;">已然</span><span style="color: black;">能够</span>用于分析X光片、CT扫描、MRI图像等,<span style="color: black;">帮忙</span>识别识别<span style="color: black;">反常</span>区域,<span style="color: black;">乃至</span>做出诊断判断。AI还<span style="color: black;">能够</span>用于识别组织切片中的细胞变异,辅助病理学家进行癌症筛查和其他<span style="color: black;">疾患</span>的诊断。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">AI还<span style="color: black;">能够</span>分析<span style="color: black;">病人</span>的基因组数据,确定最适合的治疗<span style="color: black;">方法</span>。AI<span style="color: black;">亦</span><span style="color: black;">能够</span><span style="color: black;">按照</span><span style="color: black;">病人</span>的病史和生理指标,辅助预测病情趋势。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">在<span style="color: black;">药物</span><span style="color: black;">开发</span>方面,AI<span style="color: black;">能够</span><span style="color: black;">帮忙</span>模拟化学<span style="color: black;">成份</span>的相互<span style="color: black;">功效</span>,缩短新药<span style="color: black;">开发</span>周期。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">出现</span>严重的公共卫生事件时,AI<span style="color: black;">能够</span>分析流行病数据,预测<span style="color: black;">疾患</span>传播的趋势。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在金融<span style="color: black;">行业</span>,<span style="color: black;">AI<span style="color: black;">能够</span>实时监测市场动态,识别潜在的市场<span style="color: black;">危害</span>,并制定相应的<span style="color: black;">危害</span>对冲策略。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI还<span style="color: black;">能够</span><span style="color: black;"><span style="color: black;">经过</span>分析借款人的信用记录、收入<span style="color: black;">状况</span>、消费<span style="color: black;">行径</span>等多维度数据,<span style="color: black;">评定</span>信贷<span style="color: black;">危害</span>。当然,AI<span style="color: black;">亦</span><span style="color: black;">能够</span>结合投资者的个人财务<span style="color: black;">状况</span>、<span style="color: black;">危害</span>偏好和收益<span style="color: black;">目的</span>,<span style="color: black;">供给</span>最合适的投资组合<span style="color: black;">意见</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">类似的例子实在是数<span style="color: black;">不堪</span>数。在工业制造、教育文旅、<span style="color: black;">商场</span>零售、农林牧渔、公共安全、政府治理等几乎所有<span style="color: black;">行业</span>,AI都<span style="color: black;">已然</span>有了<span style="color: black;">实质</span>的落地场景和案例。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI正在改变社会,改变<span style="color: black;">咱们</span>每一个人的工作和生活。</p>

    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><strong style="color: blue;"><span style="color: black;">█</span> <span style="color: black;">咱们</span>应该<span style="color: black;">怎样</span>看待AI?</strong></span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">AI的<span style="color: black;">商场</span>和社会价值,是毋庸置疑的。它的崛起趋势,<span style="color: black;">亦</span>是不可阻挡的。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">从企业的<span style="color: black;">方向</span><span style="color: black;">来讲</span>,AI能够<span style="color: black;">自动化重复性、繁琐的任务,</span><span style="color: black;">加强</span>生产效率和质量,<span style="color: black;">同期</span>降低生产成本和人力成本。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">针对</span>制造业和服务业<span style="color: black;">来讲</span>,这个<span style="color: black;">优良</span>至关重要,直接影响了企业的竞争力,<span style="color: black;">乃至</span>是<span style="color: black;">存活</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">从政府的<span style="color: black;">方向</span><span style="color: black;">来讲</span>,AI不仅<span style="color: black;">能够</span><span style="color: black;">提高</span>治理效率,<span style="color: black;">亦</span>能够带来新的<span style="color: black;">商场</span>模式、<span style="color: black;">制品</span>和服务,刺激经济。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">强大的AI,<span style="color: black;">亦</span>是一种国家竞争力。在科技博弈和国防事业方面,<span style="color: black;">倘若</span>AI技术不如别人,可能会带来严重后果。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">从个人的<span style="color: black;">方向</span><span style="color: black;">来讲</span>,AI<span style="color: black;">能够</span><span style="color: black;">帮忙</span><span style="color: black;">咱们</span>完成<span style="color: black;">有些</span>工作,<span style="color: black;">亦</span><span style="color: black;">能够</span><span style="color: black;">提高</span><span style="color: black;">咱们</span>的生活品质。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">从<span style="color: black;">全部</span>人类的<span style="color: black;">方向</span><span style="color: black;">来讲</span>,AI在<span style="color: black;">疾患</span>治疗、灾害预测、气候预测、消灭贫穷方面,<span style="color: black;">亦</span><span style="color: black;">能够</span>发挥重要的<span style="color: black;">功效</span>。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/cc6e57a43a134961a81d14f79f59046a~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=a%2BycZzipIIwVPB5Ijg%2FdqaRA%2Fo4%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">但事物都是有两面性的。AI<span style="color: black;">做为</span>工具,既有利,<span style="color: black;">亦</span>有<span style="color: black;">坏处</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">最现实的一个<span style="color: black;">坏处</span>,<span style="color: black;">便是</span>可能会威胁到<span style="color: black;">海量</span>的人类工作岗位,<span style="color: black;">引起</span><span style="color: black;">海量</span>失业。<span style="color: black;">按照</span>麦肯锡的<span style="color: black;">科研</span>,到2030年至2060年之间,大约50%的职业可能会逐步被AI取代,<span style="color: black;">尤其</span>是<span style="color: black;">针对</span>知识工作者而言。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/dbc9a40912694d3486bc9a68281288f6~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=Lr583wZw9BorBtDVPJAXGhgmX0g%3D" style="width: 50%; margin-bottom: 20px;">
      <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">照片</span>来自《纽约客》杂志</p>
    </div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">除此之外,AI被用于发动战争、欺诈(模仿声音或换脸,进行诈骗)、侵犯<span style="color: black;">百姓</span>权益(信息过度采集、侵犯隐私)。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">倘若</span><span style="color: black;">仅有</span><span style="color: black;">少许</span><span style="color: black;">机构</span><span style="color: black;">持有</span>先进的AI技术,可能会加剧社会的不公平现象。AI的算法偏见,<span style="color: black;">亦</span>可能<span style="color: black;">引起</span>不公平。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">AI变得越来越强大,<span style="color: black;">亦</span>会让人们产生对AI的依赖,失去独立思考和<span style="color: black;">处理</span>问题的能力。AI的强大创造力,有可能让人类失去创造的动力和<span style="color: black;">自信心</span>。</span></p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">围绕AI的发展,还有安全(数据<span style="color: black;">泄密</span>、系统崩溃)、道德伦理等一系列问题。</span></p>
    <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/tos-cn-i-6w9my0ksvp/62b2c7ee530841bca97165cb6c68bf3b~noop.image?_iz=58558&amp;from=article.pc_detail&amp;lk3s=953192f4&amp;x-expires=1727396502&amp;x-signature=9um%2FZteDUJlmxqLeFdtQBmidfqk%3D" style="width: 50%; margin-bottom: 20px;"></div>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">所有这些问题,<span style="color: black;">咱们</span><span style="color: black;">日前</span>都<span style="color: black;">无</span><span style="color: black;">可靠</span>的<span style="color: black;">处理</span><span style="color: black;">方法</span>。<span style="color: black;">因此</span>,只能在发展AI的过程中,一点点去探索、思考和<span style="color: black;">处理</span>。<span style="color: black;">针对</span>AI的<span style="color: black;">警觉</span>和防范之心,是<span style="color: black;">必定</span>要有的。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">做为</span><span style="color: black;">咱们</span>普通人,<span style="color: black;">日前</span>最现实的做法,<span style="color: black;">便是</span>先<span style="color: black;">认识</span>它、学习它。先学会<span style="color: black;">运用</span><span style="color: black;">平常</span>的AI工具和平台,<span style="color: black;">帮忙</span>自己<span style="color: black;">提高</span>工作效率,改善生活品质。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">有句话说的好:<strong style="color: blue;">“<span style="color: black;">将来</span>,淘汰你的不是AI,而是<span style="color: black;">把握</span>了AI的人”</strong>。与其焦虑,不如勇敢面对和积极拥抱,尽早<span style="color: black;">把握</span>主动权。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">好啦,以上<span style="color: black;">便是</span>今天<span style="color: black;">文案</span>的<span style="color: black;">所有</span>内容。<span style="color: black;">针对</span>一个普通人<span style="color: black;">来讲</span>,<span style="color: black;">晓得</span>这些AI常识,<span style="color: black;">便是</span>拥抱AI的<span style="color: black;">第1</span>步。<span style="color: black;">最少</span>和别人聊天的时候,谈到AI,就不会一头雾水了。</p>
    <p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;"><span style="color: black;">感谢<span style="color: black;">大众</span>的耐心阅读,<span style="color: black;">咱们</span>下期再见!</span></span></p>




4lqedz 发表于 2024-10-22 08:15:34

回顾过去一年,是艰难的一年;展望未来,是辉煌的一年。

nqkk58 发表于 2024-10-24 11:01:46

网站建设seio论坛http://www.fok120.com/

m5k1umn 发表于 2024-10-27 21:54:39

论坛的成功是建立在我们诚恳、务实、高效、创新和团结合作基础上,我们要把这种精神传递下去。
页: [1]
查看完整版本: 写给小白的AI入门科普