李开复:Google为么比BAT厉害?
<div style="color: black; text-align: left; margin-bottom: 10px;">
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa100059422efd5b988~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=SuaoddBZsCnjhub52lDxWmc63E8%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">编者按</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">创新<span style="color: black;">为何</span>看好人工智能<span style="color: black;">行业</span>?<span style="color: black;">将来</span>人类<span style="color: black;">那些</span>工作会被<span style="color: black;">设备</span>取代?<span style="color: black;">设备</span>的深度学习<span style="color: black;">为何</span>人学习不了?Google<span style="color: black;">为何</span>比BAT厉害?人工智能<span style="color: black;">行业</span>到底有<span style="color: black;">那些</span>机会?这么好的机会<span style="color: black;">怎样</span>把握?创新工场创始人李开复为您全面<span style="color: black;">诠释</span>人工智能的黄金时代。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/a9f0006f43c2da0b0fb~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=8y3qZkbxzvksuBb8o9PG%2FDg4zfU%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">为何</span>创新工场看重人工智能这个<span style="color: black;">行业</span>?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">创新工场成立于2009年,过去七年<span style="color: black;">咱们</span>专注所有中国创业者<span style="color: black;">必须</span>的服务和投资。创新工场的<span style="color: black;">特殊</span>是做技术型的投资。<span style="color: black;">咱们</span>自认为最专长的是看到一个趋势的崛起,并对这个趋势做深度的分析。<span style="color: black;">例如</span>说在2009年<span style="color: black;">咱们</span>对移动互联网的投资,在2011年<span style="color: black;">上下</span><span style="color: black;">咱们</span>对娱乐内容的投资,而<span style="color: black;">咱们</span>今天正在对人工智能做一个非常大的布局。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">为何</span><span style="color: black;">咱们</span>会<span style="color: black;">尤其</span>地看重这个<span style="color: black;">行业</span>?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">近期</span>人工智能得到<span style="color: black;">全世界</span>关注,<span style="color: black;">重点</span>是<span style="color: black;">由于</span>阿尔法狗击败了李世石。在<span style="color: black;">非常多</span>的世界<span style="color: black;">媒介</span>上<span style="color: black;">咱们</span>看到的,<span style="color: black;">便是</span><span style="color: black;">设备</span><span style="color: black;">是不是</span>会取代人脑?我个人认为<span style="color: black;">无</span>足够的科学<span style="color: black;">按照</span>,让<span style="color: black;">咱们</span>想<span style="color: black;">那样</span>多,担心<span style="color: black;">那样</span>多。而短期让<span style="color: black;">咱们</span>更应该关注的是怎么用人工智能,<span style="color: black;">咱们</span>的工具为社会创造价值,来为创业者创造机会。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">为何</span>人工智能会这么强大呢?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">便是</span><span style="color: black;">由于</span>今天人类做的大部分工作是重复性的,是基于一个有限的<span style="color: black;">行业</span>。这些工作在<span style="color: black;">将来</span>的十年都会被人工智能取代。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa1000594264216c5f9~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=I4wdMkNNdD30sQIVZY%2FkVW7qpNA%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这些工作我<span style="color: black;">这儿</span>列了<span style="color: black;">有些</span>,<span style="color: black;">倘若</span>你们不太相信的话,<span style="color: black;">咱们</span><span style="color: black;">能够</span>举几个例子:</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">今天美联社90%的稿件<span style="color: black;">已然</span>是<span style="color: black;">设备</span>人取代了,在过去的十年华尔街交易员<span style="color: black;">已然</span>一半下岗离职了。<span style="color: black;">将来</span><span style="color: black;">咱们</span>肯定会看到无人驾驶的车,当这一天来到的时候,世界上大约9%人的工作就即将被取代。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">怎么能做到呢?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">其实在这些<span style="color: black;">行业</span>里,人工智能不是模仿人类,<span style="color: black;">亦</span><span style="color: black;">不仅</span>是取代人类,<span style="color: black;">所说</span>的人工智能,<span style="color: black;">便是</span>用<span style="color: black;">尤其</span>巨大的数据集看到一个人的<span style="color: black;">一辈子</span>,<span style="color: black;">乃至</span>一百个人的<span style="color: black;">一辈子</span>,<span style="color: black;">而后</span>用数据推算出<span style="color: black;">规律</span>、判断和推测,做出比人更好的判断。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">阿尔法狗<span style="color: black;">便是</span><span style="color: black;">科研</span>了几百万的棋谱和花无数小时的时间来打败人类专家的。有人说这个不公平,一个棋手不可能花一百万个小时<span style="color: black;">科研</span>棋谱。这不是公平不公平的问题,而是<span style="color: black;">咱们</span><span style="color: black;">怎样</span>用人工智能做出人<span style="color: black;">不可</span>做的事情。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">近期</span>特斯拉提出了一个问题,<span style="color: black;">大众</span><span style="color: black;">能够</span>想想,特斯拉有几亿小时的车行数据,这些数据难道不会<span style="color: black;">作为</span>比人更好的司机吗?这一天还<span style="color: black;">无</span>到,但<span style="color: black;">倘若</span>有一个数据做十亿小时的练习,哪怕它<span style="color: black;">起始</span>像个<span style="color: black;">孩儿</span>,学完以后,难道不会超过人类吗?</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span><span style="color: black;">能够</span>想像,<span style="color: black;">倘若</span>有一个人脸识别的<span style="color: black;">设备</span>,它把全世界罪犯的脸都背下来了,有人能够做到这一点吗?<span style="color: black;">因此</span>从这个例子<span style="color: black;">能够</span>看到,<span style="color: black;">咱们</span>丝毫<span style="color: black;">不消</span><span style="color: black;">可疑</span>人工智能会全方面超越人类。<span style="color: black;">然则</span>这不是在所有的<span style="color: black;">行业</span>里,是在<span style="color: black;">有些</span>特殊<span style="color: black;">行业</span>,<span style="color: black;">或</span><span style="color: black;">能够</span>客观分析的<span style="color: black;">行业</span>,<span style="color: black;">或</span>有正确答案的<span style="color: black;">行业</span>,有大数据的<span style="color: black;">行业</span>,<span style="color: black;">况且</span>是有专家参与的<span style="color: black;">行业</span>。<span style="color: black;">因此</span>它会<span style="color: black;">逐步</span>地<span style="color: black;">出现</span>,不会一次到位的<span style="color: black;">出现</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">深度学习 巨大的变革</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">我个人有幸在三十年前<span style="color: black;">起始</span>做人工智能,但可惜当时数据<span style="color: black;">不足</span>,<span style="color: black;">因此</span>并<span style="color: black;">无</span>做太多超越人类的事情。<span style="color: black;">因此呢</span>,<span style="color: black;">咱们</span><span style="color: black;">能够</span>看到对弈只是游戏的<span style="color: black;">行业</span>,它本身并不带来巨大的<span style="color: black;">商场</span>价值。<span style="color: black;">咱们</span>在业界<span style="color: black;">更加多</span>会看到的可能是感知,<span style="color: black;">便是</span>用算法,人工智能来做判断<span style="color: black;">或</span>反馈,这个时候<span style="color: black;">设备</span>就要动起来了。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/a9f0006f43d1382e287~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=KoQDrrGfRQ1hOhsFwnHP2Ry6mxo%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span><span style="color: black;">能够</span>看到过去七八年来有<span style="color: black;">尤其</span>大的进步,这是<span style="color: black;">为何</span>呢?这是<span style="color: black;">由于</span>在<span style="color: black;">设备</span>学习的<span style="color: black;">行业</span>有一套技术,叫做深度学习。它带来了巨大的变革,今天可能<span style="color: black;">无</span>时间太解释这套算法。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">简单<span style="color: black;">来讲</span>,<span style="color: black;">便是</span>过去人是<span style="color: black;">必须</span><span style="color: black;">循序渐进</span>地告诉电脑,你该怎么走,该怎么做。对人<span style="color: black;">来讲</span>,先对人脸找眼睛、找鼻子、找耳朵,但这个大概是<span style="color: black;">所说</span>的专家系统,<span style="color: black;">便是</span>人告诉<span style="color: black;">设备</span>一切,<span style="color: black;">而后</span><span style="color: black;">设备</span>来计算,这个大概是三十年前的技术。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">大概十五年前的技术是人找特征,<span style="color: black;">设备</span>学参数。<span style="color: black;">这儿</span><span style="color: black;">便是</span>人告诉他,识别人的时候找眼睛、嘴巴,<span style="color: black;">而后</span>你再来计算,这是人机结合的学习。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><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;">深度学习的技术应该算是<span style="color: black;">近期</span>五年推出的技术,这个技术几乎人都<span style="color: black;">不消</span>参与了,<span style="color: black;">尤其</span>神奇的地方。给你看一亿张脸,你自己看,我就把<span style="color: black;">海量</span>数据给你,让你收取你认为重要的特征。<span style="color: black;">况且</span>这些特征<span style="color: black;">仅有</span><span style="color: black;">设备</span>懂,人都听不懂的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这<span style="color: black;">便是</span>阿尔法狗战李达以后世石的时候,下出了人类顶尖棋手都<span style="color: black;">无</span>办法看懂的<span style="color: black;">过程</span>。这<span style="color: black;">便是</span>深度学习,它能自我学习抽象的概念。<span style="color: black;">况且</span>这个抽象概念超越了人类的想象,<span style="color: black;">由于</span>它用的数据集比人更大。它就像一个<span style="color: black;">孩儿</span>,能够自我学习,<span style="color: black;">况且</span>数据量够大的时候,什么都学得会。这大概解释了什么是深度学习。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">深度学习,这是一个识别物体的<span style="color: black;">行业</span>,<span style="color: black;">况且</span>是有人标注的数据库。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span><span style="color: black;">能够</span>看到大概仅仅在五年前,一个<span style="color: black;">设备</span>跟人的差别还是一个74%对94%的差别。<span style="color: black;">倘若</span>算错误率的话,可能是六七倍的差别了,<span style="color: black;">因此</span>不太能够用到<span style="color: black;">实质</span>的场景里。<span style="color: black;">然则</span><span style="color: black;">每年</span>都在快速进步着,<span style="color: black;">咱们</span><span style="color: black;">能够</span>看到在2014、2015年的时候<span style="color: black;">已然</span>超越人类,<span style="color: black;">况且</span>差距越来越大。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">便是</span>说当<span style="color: black;">设备</span>识别人脸,比人更厉害的时候,<span style="color: black;">哪些</span>以识别人脸工作的人就要下岗了。<span style="color: black;">一样</span>的,语音识别超过了人,像客服,<span style="color: black;">例如</span><span style="color: black;">每日</span>打<span style="color: black;">tel</span>卖房地产<span style="color: black;">制品</span>的人,以后打<span style="color: black;">tel</span>的不是人了,<span style="color: black;">由于</span><span style="color: black;">设备</span>比人更听得懂。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">当<span style="color: black;">每一个</span>技术超越人的时候,它带来的机会是巨大的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">将来</span>五年,什么会产生巨大价值? </strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span>认为下一个<span style="color: black;">周期</span>,<span style="color: black;">将来</span>的五年什么东西会产生巨大的价值?</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">倘若</span>一个<span style="color: black;">行业</span>有<span style="color: black;">尤其</span>大的数据,只是这个数据没做挖掘,<span style="color: black;">无</span>产生价值,<span style="color: black;">例如</span><span style="color: black;">商场</span>化的流程和互联网的数据,<span style="color: black;">这儿</span>就会产生巨大的价值。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa200059bc06109274a~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=modn4fq2rdG%2BOrU%2BlS3Mq41jb2w%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">因此</span>今天谁最适合做大数据人工智能呢?其实<span style="color: black;">便是</span>BAT,在美国<span style="color: black;">便是</span>Google、微软和FaceBook,<span style="color: black;">由于</span>它们<span style="color: black;">已然</span>有了互联网大数据。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">这些怎么做到呢?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">像百度搜索的十个结果,我一个都<span style="color: black;">无</span>点,这暗示这十个都<span style="color: black;">欠好</span>,这不是绝对的暗示,<span style="color: black;">然则</span>数据足够多,是<span style="color: black;">能够</span>充分地学习。其实这些BAT都在<span style="color: black;">运用</span>,滴滴、美团都在<span style="color: black;">运用</span>,只是在创业<span style="color: black;">行业</span><span style="color: black;">大众</span>还<span style="color: black;">无</span>足够的机会。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这个世界上还是Google做得最好,<span style="color: black;">由于</span>它有世界最大的数据集,它在世界上做人工智能做得最早。<span style="color: black;">而后</span>再加上Google<span style="color: black;">数年</span>累计了<span style="color: black;">海量</span>的技术人才,还有<span style="color: black;">尤其</span>快的<span style="color: black;">设备</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这<span style="color: black;">便是</span><span style="color: black;">为何</span>Google<span style="color: black;">能够</span>做出来阿尔法狗,而不是BAT、美团、小米,<span style="color: black;">然则</span>Google的野心<span style="color: black;">不仅</span>是阿尔法狗。前些日子,<span style="color: black;">咱们</span>看到Google自我命名为Alphabet,是互联网的一家<span style="color: black;">机构</span>,它可能还有做医学的<span style="color: black;">机构</span>,可能还要做基因检测的<span style="color: black;">机构</span>,可能做汽车等等<span style="color: black;">行业</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">Google做智能<span style="color: black;">制品</span>的过程中发展了一套Google大脑。其实Google里有两套大脑,这个<span style="color: black;">能够</span>用到各个<span style="color: black;">行业</span>。<span style="color: black;">发掘</span>这个大脑,其实才是它的最大价值。在其它的<span style="color: black;">行业</span>,只要有数据就<span style="color: black;">能够</span><span style="color: black;">起步</span>。<span style="color: black;">因此</span>Google<span style="color: black;">持有</span>了巨大的数据量,它们启动了数十个<span style="color: black;">区别</span>的项目。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在美国有<span style="color: black;">有些</span><span style="color: black;">公司</span>,<span style="color: black;">例如</span>OpenAI,<span style="color: black;">亦</span>担心<span style="color: black;">倘若</span>Google这么<span style="color: black;">都数</span>据,还有<span style="color: black;">无</span>机会跟它竞争,它会不会比人类<span style="color: black;">晓得</span>的太多了?当然最好的办法是让<span style="color: black;">更加多</span>的人参与到这个<span style="color: black;">行业</span>来创业,<span style="color: black;">这般</span>就不会被某一个垄断了。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">怎样</span><span style="color: black;">捉捕</span>这个<span style="color: black;">行业</span>的机会?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这么好的机会,<span style="color: black;">咱们</span>怎么<span style="color: black;">捉捕</span>呢?我告诉<span style="color: black;">大众</span>,这并不是人人能做,做起来<span style="color: black;">亦</span>不是很简单的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">1有些<span style="color: black;">行业</span>是要平台</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">例如</span><span style="color: black;">研发</span>一个苹果APP,<span style="color: black;">或</span>安卓的APP,<span style="color: black;">必要</span>要有平台的。计算机程序员只<span style="color: black;">能够</span>学会<span style="color: black;">研发</span>一个安卓APP,即便是一个优秀的程序员<span style="color: black;">亦</span>没法儿用深度学习,<span style="color: black;">由于</span>这个还<span style="color: black;">无</span>平台化,你还要<span style="color: black;">晓得</span><span style="color: black;">非常多</span>微妙的东西,这方面的专家可能<span style="color: black;">仅有</span>几百个人,<span style="color: black;">因此</span><span style="color: black;">这般</span>的人才很稀缺。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">2<span style="color: black;">必须</span>海量的数据</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">说一千万的数据<span style="color: black;">便是</span>大数据,其实那是小数据。真的要深度学习有用武之地,<span style="color: black;">必须</span>有更巨大的量。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">3数据储存,还有存储量4<span style="color: black;">咱们</span>都认为深度学习超越人了,<span style="color: black;">能够</span>倒过来解释给<span style="color: black;">咱们</span>听,<span style="color: black;">然则</span><span style="color: black;">设备</span>学习<span style="color: black;">无</span>办法告诉你,它怎么决定的,你要相信就相信,<span style="color: black;">根据</span>结果说话。你让阿尔法狗回<span style="color: black;">来讲</span>怎么打赢李世石的,它说不出来。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">怎么做呢?买这个数据,标注这个数据,不要浪费钱,GPU可能比CPU好用。找专家,<span style="color: black;">这般</span>的专家真的不多。Google<span style="color: black;">已然</span>开出九位数的代价挖<span style="color: black;">这般</span>的专家。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这是<span style="color: black;">咱们</span>过去<span style="color: black;">不可</span>够想象的,在国内<span style="color: black;">咱们</span><span style="color: black;">亦</span>看到,<span style="color: black;">例如</span>说看到投资的Face++,<span style="color: black;">近期</span><span style="color: black;">亦</span>是<span style="color: black;">奋斗</span>挖到了华人深度学习的<span style="color: black;">科研</span>员,这些人才的价值<span style="color: black;">尤其</span>巨大的,<span style="color: black;">由于</span>非常稀缺。<span style="color: black;">然则</span>话说回来,这个稀缺<span style="color: black;">亦</span>是<span style="color: black;">将来</span>两三年的现象,之后我觉得会有<span style="color: black;">非常多</span>聪明人,尤其在中国会涌入这个<span style="color: black;">行业</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">有这些东西还<span style="color: black;">不足</span>,深度学习还会有<span style="color: black;">有些</span>挑战,最后还<span style="color: black;">不可</span>完成你的应用,该怎么办呢?就要用巧妙的、非技术的<span style="color: black;">办法</span>,不要取代人,辅助人就好了。<span style="color: black;">例如</span>做安防的人,他只要带着摄像头,摄像头把拍出来的东西告诉人,那边的窗子坏了,其实就够了,辅助人就够了。还有辅助医学,<span style="color: black;">咱们</span>不是说取代<span style="color: black;">大夫</span>,<span style="color: black;">然则</span><span style="color: black;">能够</span>给<span style="color: black;">大夫</span>做<span style="color: black;">有些</span>工具,这是完全<span style="color: black;">能够</span>的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">咱们</span>是不是能局限于某个<span style="color: black;">行业</span>或场景?</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">例如</span>,<span style="color: black;">为何</span>要全天候无人驾驶?<span style="color: black;">为何</span>在可控的<span style="color: black;">行业</span>做好无人驾驶呢?<span style="color: black;">此刻</span>特斯拉的车,<span style="color: black;">咱们</span>看到它这次出车祸(笔记侠注:6月末,特斯拉首起死亡事故被<span style="color: black;">揭发</span>。一辆自动驾驶的Model S拦腰撞上一辆正在横贯马路的货车,<span style="color: black;">导致</span>车主死亡),<span style="color: black;">便是</span>分不清巨大的卡车和天空的颜色。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">因此</span>,要用一个聪明的局限<span style="color: black;">行业</span>来创造价值。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa200059bbd433715c5~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=LE0Ta%2BZRUVur0PokzU0Q%2BRmEEdM%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">这是今天一张重要的PPT,<span style="color: black;">然则</span>我<span style="color: black;">亦</span><span style="color: black;">不可</span><span style="color: black;">保准</span>这张是<span style="color: black;">无</span>错的,<span style="color: black;">由于</span><span style="color: black;">无</span>人<span style="color: black;">能够</span>预测<span style="color: black;">将来</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">咱们</span>相当看好大数据,尤其是金融<span style="color: black;">行业</span>,<span style="color: black;">由于</span>它有最大的黑盒(笔记侠注:<span style="color: black;">由于</span>人大脑太<span style="color: black;">繁杂</span>,<span style="color: black;">因此</span>干脆把他看作一个“黑盒”),能够产生最大的效应,<span style="color: black;">亦</span>能吸引最大的应用;在感知方面有一个误解,<span style="color: black;">便是</span>自然语言理解,以为听到语音了,变成字了,我就懂了,<span style="color: black;">由于</span>人是很容易懂的。<span style="color: black;">然则</span>听、识别你讲的字,和懂你的意思还是有一个<span style="color: black;">尤其</span>大的鸿沟。这点我觉得<span style="color: black;">最少</span>五年,<span style="color: black;">乃至</span>十年的时间<span style="color: black;">才可</span>突破。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">设备</span>人方面,<span style="color: black;">大众</span><span style="color: black;">皆想</span>做家庭<span style="color: black;">设备</span>人,但<span style="color: black;">这般</span>是<span style="color: black;">不可</span>满足家庭期望值的。这个在日本可能是<span style="color: black;">能够</span>的,<span style="color: black;">然则</span>我不认为有<span style="color: black;">必定</span>的<span style="color: black;">表率</span>性。<span style="color: black;">这般</span>的<span style="color: black;">设备</span>人好玩儿,<span style="color: black;">能够</span>做老人<span style="color: black;">陪同</span>,<span style="color: black;">能够</span>做小孩的玩偶,但<span style="color: black;">便是</span>这么多了。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">你要期待它帮你炒菜、扫地,<span style="color: black;">作为</span>家庭主妇,<span style="color: black;">或</span><span style="color: black;">作为</span>家庭新的一员<span style="color: black;">或</span>新的电器,这个还是一个天方夜谭。当然,你很聪明地把这个<span style="color: black;">设备</span>人做<span style="color: black;">成为了</span>一个音响,<span style="color: black;">或</span>做<span style="color: black;">成为了</span><span style="color: black;">陪同</span>老人的小<span style="color: black;">设备</span>,像<span style="color: black;">咱们</span>投资的小鱼在家,当然<span style="color: black;">亦</span>是<span style="color: black;">能够</span>做的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">然则</span>,<span style="color: black;">咱们</span>人所想象的<span style="color: black;">设备</span>人<span style="color: black;">必定</span>会先在<span style="color: black;">商场</span><span style="color: black;">行业</span><span style="color: black;">显现</span>,之后才会在家庭有<span style="color: black;">期盼</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">最后关于无人驾驶,让它在有些场景<span style="color: black;">能够</span>有用,先<span style="color: black;">帮忙</span>人,人<span style="color: black;">帮忙</span>它,最后才是无人,而这还是在十年以后。总之,这个<span style="color: black;">行业</span><span style="color: black;">咱们</span>是<span style="color: black;">尤其</span>看好,<span style="color: black;">期盼</span>今天的演讲对<span style="color: black;">大众</span>有些启发。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">谢谢。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p26-sign.toutiaoimg.com/aa200059bbf2e89ce8a~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=JGStQOYPn60jbkVCEd3gdMsHJUo%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">延 伸 阅 读</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa100059427fb5619df~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=hjrL9LDwsfov9%2BClLzX7tT%2FUMyA%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">李开复:资本寒冬终将过去 人工智能时代来临</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">创新工场创始人李开复出席2016硅谷高科技创新·创业峰会(北京)暨<span style="color: black;">全世界</span>产业互联网大会,<span style="color: black;">针对</span>人工智能主题进行了主旨演讲。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">智能时代到来:基于客观思考的工作终究被人工智能取代</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">李开复指出人工智能的强大,是<span style="color: black;">由于</span>今天人类做的大部分工作是重复性的,是基于<span style="color: black;">能够</span>客观思考<span style="color: black;">乃至</span>穷举的这些工作在<span style="color: black;">将来</span>的十年都会被人工智能取代,<span style="color: black;">例如</span>今天美国的美联社90%的稿件<span style="color: black;">已然</span>是<span style="color: black;">设备</span>人取代,在过去的十年华尔街交易员<span style="color: black;">已然</span>一半下岗离职。当这一天来到的时候,世界上大约9%的人<span style="color: black;">她们</span>部分<span style="color: black;">或</span><span style="color: black;">所有</span>的工作就即将被取代。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">深度学习:智能<span style="color: black;">设备</span>超越人类想象</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">他<span style="color: black;">暗示</span>人工智能在过去并<span style="color: black;">无</span>做太多超越人类的事情,发展对弈只是游戏的<span style="color: black;">行业</span>,本身并不带来巨大的<span style="color: black;">商场</span>价值,但<span style="color: black;">此刻</span>却迎来巨大变革,这是<span style="color: black;">由于</span>在<span style="color: black;">此刻</span><span style="color: black;">设备</span>学习的<span style="color: black;">行业</span>有一套技术叫做深度学习。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/a9f0006f43eb12a0021~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=JCQNkqizeHNujodjqrWJTuvv9%2FI%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">简单<span style="color: black;">来讲</span>,深度学习的技术<span style="color: black;">此刻</span>几乎人都不参与,只是把<span style="color: black;">海量</span>数据传给<span style="color: black;">设备</span>,<span style="color: black;">设备</span>负责收取认为重要的特征。这<span style="color: black;">便是</span>阿尔法狗战李达以后世石时,下出了人类顶尖棋手都<span style="color: black;">无</span>办法看懂的<span style="color: black;">过程</span>。这<span style="color: black;">便是</span>深度学习,它能自我学习,<span style="color: black;">况且</span>这个抽象概念超越了人类的想象。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">AI<span style="color: black;">行业</span>:技术仍需改进 VC投资需<span style="color: black;">小心</span></strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa200059bc252e4f14d~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=zVjXid0VA0vCJWH%2FE7iGNlD24J8%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">投资方面,李开复谈到人工智能确实是<span style="color: black;">此刻</span>很热的<span style="color: black;">行业</span>,但其实当下<span style="color: black;">无</span><span style="color: black;">那样</span>多<span style="color: black;">机构</span><span style="color: black;">能够</span>投资,在国内<span style="color: black;">乃至</span><span style="color: black;">全世界</span>,人工智能的专家并不多,容易有<span style="color: black;">泡泡</span>市场。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">并且AI<span style="color: black;">日前</span>还局限于一个<span style="color: black;">行业</span>和场景,在技术方面仍有很长的路要走。<span style="color: black;">例如</span>现特斯拉的车,这次出车祸,<span style="color: black;">便是</span><span style="color: black;">由于</span>它分不清巨大的卡车和天空的颜色,<span style="color: black;">必须</span>找一个<span style="color: black;">行业</span>不会有巨大卡车,<span style="color: black;">况且</span>阿尔法狗怕的天黑、下大雪,就需找一个<span style="color: black;">区别</span>的场景,<span style="color: black;">因此</span><span style="color: black;">亦</span>要寻求一个聪明的局限<span style="color: black;">行业</span>来创造价值。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">资本寒冬:波动正常 保持核心技术是关键</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">自2015年<span style="color: black;">败兴</span>,股市<span style="color: black;">连续</span>动荡,二级市场<span style="color: black;">连续</span>经济低糜,给创业市场泼了一场冷水,创业者开始高呼,资本寒冬已来临。对此李开复谈到,一级市场和二级市场<span style="color: black;">同样</span>都有波动,<span style="color: black;">此刻</span>的状态可能不大<span style="color: black;">阳光</span>,但<span style="color: black;">长时间</span>来看肯定还是涨幅态势,从过去两千年到<span style="color: black;">此刻</span>,跌宕周期<span style="color: black;">非常多</span>次,但<span style="color: black;">长时间</span>还是呈<span style="color: black;">提升</span>的趋势。AR<span style="color: black;">或</span>VR,<span style="color: black;">将来</span>的前景还是<span style="color: black;">非常</span>巨大。短期内,<span style="color: black;">有些</span><span style="color: black;">机构</span>可能会估值过高,<span style="color: black;">或</span>整体有融资方面的挑战,这个是创业<span style="color: black;">时期</span>毋庸置疑会经历的过程。<span style="color: black;">针对</span>一个投资<span style="color: black;">机构</span><span style="color: black;">来讲</span>,<span style="color: black;">亦</span>许会经历短期平缓的时间,<span style="color: black;">乃至</span>有可能下降,但它整体还是成长的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa10005942c3d8f6268~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=olbSV05o30KdZ2npx6KJ85pLsYA%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">相应的在资本寒冬下,创业<span style="color: black;">机构</span>倒闭成风,尤其是O2O型的创业<span style="color: black;">机构</span>,虽然很受资本市场欢迎但行业内的竞争过于同质化,<span style="color: black;">更加多</span>的依靠烧钱进行扩张,<span style="color: black;">泡泡</span>市场严重,2016年资本收缩,<span style="color: black;">非常多</span>创业<span style="color: black;">机构</span>都宣布告终。但<span style="color: black;">针对</span>有些创业<span style="color: black;">机构</span>热潮中仍旧能存活下来,且依旧在蓬勃发展的<span style="color: black;">机构</span><span style="color: black;">一起</span>点,李开复指出这些<span style="color: black;">机构</span><span style="color: black;">常常</span>核心技术很好,<span style="color: black;">或</span>有不错的市场核心<span style="color: black;">优良</span>,在<span style="color: black;">这般</span>的一个状态下会有很好的成长速度与成长方式。但他<span style="color: black;">亦</span><span style="color: black;">这里</span>提醒广大的从创业者,一方面要<span style="color: black;">保证</span>自己走在最前面不会保持<span style="color: black;">优良</span>状态,<span style="color: black;">另一</span>一方面,<span style="color: black;">必定</span>要节省<span style="color: black;">而后</span>活下来,<span style="color: black;">这般</span>不会受到VC的欺负,并懂得与创造收入,有足够的收入活下来,这<span style="color: black;">亦</span>是<span style="color: black;">非常</span>重要的。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">大学生创业:不急于一时 多<span style="color: black;">累积</span>经验</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在资本的鼓励下,不少大学生<span style="color: black;">亦</span><span style="color: black;">起始</span>创业,但<span style="color: black;">亦</span>在资本市场的洗礼下备受争议。<span style="color: black;">例如</span><span style="color: black;">近期</span>被议论纷纷的未满十八岁的创业者—神奇百货CEO王凯歆事件,使得<span style="color: black;">非常多</span>人对<span style="color: black;">青年</span>大学生的创业热潮并不看好。对大学生创业,李开复<span style="color: black;">意见</span>想创业的大学生,<span style="color: black;">倘若</span><span style="color: black;">无</span>过创业经验,最好要去创业<span style="color: black;">机构</span>学习,并要找到好的切入点,技术型的海归创业者做技术型<span style="color: black;">机构</span>更有<span style="color: black;">优良</span>,多向国外学习与吸取经验,<span style="color: black;">日前</span>中国的市场<span style="color: black;">非常</span>良好,<span style="color: black;">能够</span>更好更快的规模化。中国创业<span style="color: black;">机构</span><span style="color: black;">亦</span>要借鉴国外的创业方式,<span style="color: black;">例如</span>中国的创始人有着较高的决策权,由最高职位决策者来决定<span style="color: black;">机构</span>的发展,而美国的创业<span style="color: black;">更加多</span>的是<span style="color: black;">一块</span>共事,不急于一时的利益。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">源自</span>丨笔记侠 T3plus</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">本文责任编辑丨Chaona</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="https://p3-sign.toutiaoimg.com/aa10005942d9084da46~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723729012&x-signature=7XO1eX2Pw%2FhKTA6rkyrpsUXMiUg%3D" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">更加多</span>精彩内容,请搜索关注<span style="color: black;">微X</span>公众账号:“<span style="color: black;">全世界</span>创新论坛”或“bdqqcxlt”</p>
</div>
同意、说得对、没错、我也是这么想的等。 我完全同意你的观点,说得太对了。
页:
[1]