大会演讲·中文精读

马斯克亮相 2023 迪拜世界政府峰会

Elon Musk at the 2023 World Government Summit in Dubai

中文精读由 AI 基于原始视频与文稿改写,非官方翻译,可能存在偏差或错误;表述以原视频为准。说明

迪拜世界政府峰会的台上,坐着 150 多个国家的政府代表。主持人问马斯克:你花了四百四十亿美元买下 Twitter,为什么不干脆自己做一个新平台,说不定还更便宜?

马斯克说他想过从零开始,但自己做一个大概要三到五年,而买下现成的能省下这段时间。在他看来,现在这个阶段的三到五年,值很多钱。

这场对话持续了将近一个小时,从 Twitter 一直聊到 AI 监管、教育、外星人和他每天睡几个小时。最有意思的部分,是他当着满屋子的政府官员说了一句可能让主办方有点尴尬的话:你们合作得太紧密,本身也是一种风险。

从零做一个社交平台要三到五年,他等不起

主持人的追问很直接:自己做不是更省钱吗?

马斯克的回答绕开了钱。他说技术正在加速,加速度本身有价格。三到五年在别的时代可能不算什么,但在 2023 年,三到五年意味着错过一整代技术迭代。

另一层原因是他对社交媒体方向的担忧。他想看到一个最大程度被信任的数字公共广场(digital public square):国内的人能在上面说话,跨国的人也能在上面说话,审查降到法律允许的最低线。

他特意强调了"法律允许"这四个字。不同司法辖区的法律差别很大,但总体原则应该是社交媒体公司遵守所在国的法律,而不是在法律之外再伸一只手去拨天平。

"社交媒体公司应该遵守各国的法律,而不是试图在各国法律之外去动天平。这个想法,多数国家的立法者和民众大概都会认同。"(马斯克)

他把那只"多余的手"具体指了出来:旧金山和伯克利的那套意识形态。他用了一个词,niche,小众的。在他看来,那套价值观在全世界范围内属于小众,但 Twitter 过去在把它往全球推。

"我认为这对文明的未来很重要,要把那只压在天平上的手纠正过来,让 Twitter 更准确地反映地球上人们的价值观。"

X 的终局:一个能付钱、能通信、能找到真相的万能应用

主持人问五年后 Twitter 是什么样子。

马斯克给出的答案不是 Twitter,是一个很老的名字:x.com。那是他 1999 年创办的在线银行,后来合并成了 PayPal。他想把这个名字重新捡起来,做成一个"万能应用"(everything app)。

他列了功能清单:支付、金融服务、信息流、安全通信,任何数字化的东西。核心是"最大程度有用"和"最大程度有趣"。

他还给了这个应用一个更有野心的定位:真相的来源。你想知道世界上到底在发生什么,打开 X,就能查到。

在他的描述里,Twitter 是这个终局的加速器,不是一个终点。

用"不后悔的分钟数"衡量产品

主持人问得很尖锐:Twitter 上有很多错误信息,不同族群、不同国家的人在上面互相攻击,这个怎么修?

马斯克没有回避,他先讲了一个正在推的东西,社区笔记(Community Notes)。目前只支持英文,之后会扩展到所有语言。

他觉得这套机制好在,它把"判断真假"变成了一场竞争。不是投票,是竞争:谁提供的信息最准确,谁就能被顶上去。

"这是一种对真相的竞争,我认为这个概念非常有力。"(马斯克)

主持人追问:真相这种东西,一部分人认为真的,另一部分人可能不认。马斯克的回应是,那就要一个最接近的近似值。

第二件事是身份层。他希望尽可能多的组织、人物、机构通过验证,并且把组织归属标清楚。一个账号到底属于某位议员、某位记者,还是迪士尼公司,打开就能确认。

他的推理是:一旦这些身份互相咬合起来,欺骗就变得很难,而且账号背后的人有声誉要维护,说话自然会收敛。

然后是这场对话里最值得记下来的一个指标:不后悔的使用时长(unregretted user minutes)。

他说衡量产品不能只看总使用时长,要看用户用完之后不后悔的那部分。他举了 TikTok 做对比:经常听到有人说自己在 TikTok 上花了两个小时,但后悔了那两个小时。他补了一句,不是要踩 TikTok,只是不希望 Twitter 变成那样。

"如果你在 Twitter 上花了半小时,觉得有用、有趣、对你的生活是件好事,最终对整个文明也是一股向善的力量。这是我们的期望。"(马斯克)

他给在场政府官员的建议:自己发推,别找代笔

主持人问:现场有 150 多个政府,管着 80 亿公民,政府该怎么用 Twitter 服务民众?

马斯克的建议出奇地朴素:多说话,用自己的声音说话,别用你以为自己"应该"用的那种声音说话。

他说很多人的问题在于,一开口就变成了另一种腔调,听起来僵硬,不真实。他打了个比方,企业发的新闻稿,读起来就像宣传材料。

所以他建议 CEO、部长、议员都真实地说话。有具体政策,就把它讲清楚。

有人担心挨批评。"挨点批评真没什么大不了。"他自己在 Twitter 上长期被攻击,他说他不在意,只是这行需要脸皮厚一点,因为有些人真的是拿刀往里拧。

他特别反对找人代管账号。有些人会请一个"Twitter 经理"来替自己发,他觉得没这个必要。推文应该自己写,写错了也没关系。就像你自己上台演讲、自己开会一样,亲自发,把想传达的信息直接传出去。

全场最不合时宜的一句话:政府之间合作太多也是风险

说完这些,马斯克自己加了一段。他说他知道这个会叫"世界政府峰会",但他觉得大家应该对"真的变成一个单一世界政府"这件事保持一点警惕。

"坦白说,这可能听起来有点怪,政府之间合作太多,反而会制造文明层面的风险。"(马斯克)

他给了历史依据。文明有兴有衰,但人类整体没有因此灭亡,原因是这些文明彼此被遥远的距离隔开。罗马在衰落的时候,伊斯兰世界正在兴起,哈里发国家干得非常好,而罗马干得非常糟。那段时间,知识反而被保存下来了,科学也有大量进展。

他的推论是:如果全人类压缩成一个文明,那这个东西一旦崩,就整体崩了。他不主张打仗,只是希望保留一些文明多样性,某个部分出问题的时候,整体不至于一起塌掉。

主持人现场回应了这个观点,既有认同也有反驳。他指出今天的人不再用刀剑打仗,手里是核武器,真打起来整个人类文明都没了。而他所在的阿联酋正在做的事,恰好是另一个方向的实验:180 个国籍、各个种族、各种宗教的人住在一起,试图证明人类可以和平共处。

马斯克的回应只有一句:那挺好的。这一段没有更多交锋。

推特 CEO 什么时候招:等公司站稳

主持人追问他一个人兼任董事长、所有者、CEO,什么时候找个接班的。

马斯克说现在的任务是稳住组织:让财务状况健康,把产品路线图清晰地铺出来。他估计大概在今年年底,那时候公司应该处于一个稳定的位置,适合交给别人来管。

十年技术预测:能源转型慢得像换车队,AI 快得让人担心

主持人问他未来十年的技术走向。马斯克先说了句实话:十年这个尺度上的技术预测很难精准,变化太快。

第一条线是可持续能源:太阳能、风能、电池、电动车。这条线的增长率很高,但基数太大。他给了一个很能说明问题的数字:就算明天开始所有新车都是电动的,把路上的存量车全换掉也要二十年。整个转型是三十到四十年的量级,慢。

第二条线近得多,也急得多:人工智能。

他说 ChatGPT 让人们第一次看清 AI 已经走到哪一步了。但技术本身并不新,先进状态已经持续了好几年,缺的只是一个普通人能用的界面。ChatGPT 做的事,是给一项存在已久的技术装上了一层可访问的界面。而更先进的版本还在后面。

所以他给出一个明确主张:AI 需要被监管。

OpenAI 是怎么来的,以及他现在的立场

他顺手把 OpenAI 的来历讲清楚了。

当年他担心 Google 对 AI 安全不够上心,于是和几个人一起创办了 OpenAI。最初是非营利、开源的。现在它变成了闭源、营利。他把自己和这家公司的关系说得很干净:没有股份,不在董事会,也不以任何方式控制它。

这段话放在 2023 年初,其实是给后面一连串争议提前划清界限。

为什么 AI 该像飞机和药物一样被监管

马斯克的论证方式很工整。凡是可能对公众构成风险的技术,飞机、汽车、药品,都有监管机构盯着。AI 的风险比这三样都大,那按理也该有类似的监管。

他也承认监管会让 AI 慢一点,但他觉得慢一点可能是好事。

麻烦在于监管的节奏。政府监管机构通常是"出了事才建"的。汽车和飞机早期都没有监管,代价是一堆事故和死亡,有些厂商还在偷工减料,公众受不了了才建起机构。今天民航极其安全,每英里的死亡率比开车还低,汽车本身也比过去安全得多。

他举了安全带。汽车行业为了抵制强制安全带,抗争了十年到十五年,最后监管者强推,安全水平大幅提升。安全气囊是另一次大改进。

他担心的是:AI 万一出事,监管的反应速度可能来不及。

"这可能是文明未来最大的风险之一。它有巨大的前景,巨大的能力,但与之相伴的是巨大的危险。就像核物理,你得到了核电,也得到了核弹。"(马斯克)

教育的顺序错了:先讲问题,再讲工具

话题转到教育,马斯克的表达明显更兴奋。

他先讲了一个关于记忆的判断:人的大脑演化出来是为了遗忘,凡是它判断为不重要的东西,都会被丢掉。他拿手机做对比:手机能把整本百科全书记到最后一个字母、最后一个像素,人的记忆力在它面前不值一提。所以大脑一直在做的事,其实是删东西。

由此推出他的第一条教育原则:必须告诉孩子为什么要学这门课。有了相关性,才有动机。

第二条原则更具体:把知识当作问题的解来教。

他的例子是拆一台内燃机。要拆它,你需要扳手、螺丝刀和一堆工具。拆的过程中,你自然理解了这些工具为什么存在。数学、物理、工程都是工具,如果先有问题,工具的意义就建立起来了,记起来也容易得多,因为它们能解释世界是怎么运转的。

现在的教法反过来了:直接教工具,不讲问题。他打了个很刻薄的比方,那等于开一门"螺丝刀课"、一门"扳手课",上完了你也不知道为什么会有这门课。

他承认有些课的存在本身就有问题,因为"为什么要学"这个问题根本给不出好答案。他点名了高等数学:大部分人不会觉得它有用,一辈子也用不上;真正用得上的那部分,作为一般原则很快就能讲完。

第三条原则是批判性思维,而且要从很小的年纪开始教。他把它比作一道心理防火墙。别人告诉你一件事,你要能判断:它站得住吗?它有多大概率是真的?然后据此拒绝更可能是假的东西,接受更可能是真的东西。

十二年和四年,可能都长了

主持人问:十二年学制,你觉得对吗?

马斯克的回答很温和。他说人成熟本来就需要时间,情绪成熟、身体成熟、心智成熟和教育是同时发生的。他猜十年可能也够,不一定非要十二年。但十六岁成熟的人比十八岁的少,所以十二年大概不算差。

然后他补了一刀:大学那四到六年,可能就有点过头了。

"我们大概可以砍掉几年,问题也不大。"(马斯克)

他的孩子被 Reddit 和 YouTube"编程"了

主持人问他自己对孩子刷社交媒体有没有规矩。

马斯克承认他基本没怎么限制,然后自己评价了一句:这可能是个错误。他说几个孩子主要被 Reddit 和 YouTube 影响,比别的东西都厉害。

他建议家长做两件事:限制得比自己过去更严一点,并且搞清楚孩子到底在看什么。

他的说法很有意思:孩子们正在被某个社交媒体算法"编程",而这个算法是硅谷写的,你未必认同它。所以家长要监督,要警惕。

一周八十小时,是他现在的"理想状态"

主持人提到他一天干二十个小时,睡办公室沙发。

马斯克纠正了一下:二十小时的工作日非常少见,而且相当痛苦。他每天睡六小时。经验是睡不够六小时,醒着的时间更长,但干的活更少。

他承认自己工作时长相对多数人来说不合理,基本一周七天,从睁眼干到闭眼。他强调这不是给所有人的建议,他自己其实想少干一点。

为什么之前必须这么拼:特斯拉经历过一段贴着悬崖边飞的日子,差点破产,他不把所有东西投进去,公司就没了。

现在不一样了。运营特斯拉需要的工作量比 2017 到 2019 年少得多,公司有了规模效应,不再是生死问题。SpaceX 有很强的团队,他少去也能推进。Twitter 则还是一个"反向创业公司",需要大量工作才能稳定下来,把软件工程的引擎建起来,把产品路线图和能执行路线图的人配齐。

他给了一个数字作为目标:一周八十小时就很舒服了。

不明飞行物不是外星人,但没有外星人才更让人不安

最后一个问题,主持人问阿拉斯加、休伦湖和加拿大上空被打下来的三个飞行物,是不是外星人。

马斯克说不是。

然后他聊到了费米悖论:宇宙的年龄有 138 亿年,如果真的这么老,外星人应该到处都是,可它们在哪儿?他说自己没见过任何外星技术或外星生命的证据。他补了一句,SpaceX 做了很多事,论空间技术,他大概知道得不比任何人少。

让他不安的是另一种可能:如果确实没有外星人,那意味着意识和文明在宇宙里就像一根巨大的黑暗中的小蜡烛,脆弱到随时可能被吹灭。

"我们应该格外小心地对待这根可能存在的、黑暗中的小蜡烛,确保它不熄灭,把意识之光延伸到地球之外,尽一切努力让它继续亮着。"(马斯克)

整场对话里,这段话最不像商业演讲,也最像他真正在意的那个问题。买 Twitter、造火箭、做电动车,在他的叙事里都指向同一件事:让这根蜡烛别灭。

这场对话能带走什么

  • 评估一个内容产品,别只看时长。看"不后悔的时长":用户用完之后是觉得赚了还是亏了。做产品的人和刷产品的人都该问这一句。
  • 看平台上的信息真假,一条可操作的标准是:看有没有公开的、可竞争的纠错机制。能被人当场纠正并留下痕迹,比任何官方声明都可靠。
  • 如果你在机构里负责对外发声,自己写,别外包。真实的语气比正确的措辞更有说服力,写错了道个歉就行,代笔写出来的东西没人信。
  • 判断一项技术的风险,可以套马斯克那个问法:它对公众的风险比飞机、汽车、药品更大吗?如果是,它迟早会有监管,早关注比晚关注好。
  • 教孩子任何东西之前,先回答"为什么要学"。大脑只保留它认为重要的东西,讲不清用途的课程,讲完就会被忘掉。
  • 把孩子从算法手里抢回来一点:限制时长,更重要的是知道他们在看什么。推荐算法是有立场的,只是它不告诉你。
  • 关于工作时长,他自己给出的数字是每天六小时睡眠、一周八十小时为目标。睡不够的时候,醒着的时间更长,产出反而更少。这条比工作时长本身更值得记住。
  • 最后一条,来自他对文明多样性的担忧:任何系统,包括你的团队和你的技能组合,都需要保留一点冗余。全部押在一个方向上,效率高,但没有备份。
YouTube 平台字幕(自动或人工),可能有识别错误,人名与数字请以视频为准。已按语句合并成段落,并去掉 um、uh 等口头填充词。 文稿来源:YouTube(自动/人工字幕)。

0:00hello good morning good morning good to see you alone it's been almost six years where we I sat with you here in this platform with a great audience it was your first trip to Dubai with your family I hope you enjoyed it yeah it was wonderful I'm very much into it and I I'm I see my my head is Gigantic on this day

0:31my head has grown larger since we last met is it is it is it because of Twitter yeah I don't know perhaps it Twitter is it's certainly quite the roller coaster Elon just you know it's been it's been six years within six feet we've seen a tremendous thing since our our last conversation we've seen the pandemic Russian Ukrainian War

1:01development of chat gbt you launched Starship you recently also acquired Twitter can I ask you this question why why why you bought Twitter why didn't you create your own platform maybe it was cheaper for you I mean I thought about creating something from scratch but I I thought it would Twitter would perhaps accelerate progress versus creating something from scratch by three to five years and I think we are seeing just a tremendous technology acceleration that you know three to five years is actually worth a lot so I mean if you think I was a little worried about the direction that and the effect of social media on the world and especially Twitter and

2:03I thought it was very important for there to be a maximally trusted sort of digital Public Square where people you know within countries and internationally could communicate with you know the least amount of censorship Allowed by law obviously that varies a lot by jurisdiction but I think in general you know social media companies should adhere to the laws of of countries and not try to put a thumb on the scale beyond the laws of countries so and I think this is something that is probably agreeable to the you know the legislators and and the people of most countries so so I think it's that's the general idea is just to reflect the values of of the people

2:59as opposed to imposing the values of essentially San Francisco and Berkeley which are somewhat of a niche ideology as compared to the rest of the world and but but you know Twitter was I think doing a little too much to impose a niche as a it's near San Francisco Berkeley ideology on the world so you know I I thought the it was important for the future of civilization to try to correct that thumb on the scale if you will and and and just have Twitter more accurately reflect like I said the the values of the the people of Earth

3:49that's that's the that's the intention and hopefully as we succeed in doing that yeah but how do you see Twitter if we we say it five years down the road what's your vision for for this platform what what should it do well I think it'd be I'd like to you know have the sort of long-term vision for something called x.com from back way back in the day which is kind of like a sort of like an everything app where it's just massively useful it does you know payments does so it provides Financial Services provides information flow really anything digital and it also provides secure Communications so it relates to

4:49you know I think it would be as useful as possible as entertaining as possible and thank you I went to find out what's going on and what's really going on then you could should be able to go on on you know X the X app and and find out so it's a sort of source a sort of a source of Truth and a maximally useful I guess app is about the wrong word but system and and Twitter is essentially an accelerant to that sort of maximally useful everything app

5:32yeah how how you are gonna I mean if you look at Twitter today I mean it's it's a platform sometimes there is a lot of misinformation in Twitter sometimes I don't feel comfortable even because there is some way there is this negative between Nation between people between different ethnic group that is the same thing how we are how we are going to fix this issue where you are you are in a mission with for Humanity to get them together

6:06yeah I think there's there's something that we're putting a lot of efforts into called Community notes it's currently just in English but we will be expanding it to all languages that is I think quite a good way to assess the truth of things where it's the community itself basically the you know the people of Earth who are basically what you know not exactly voting but but competing to provide the most accurate information so it's sort of a competition for truth and I think it's a very powerful concept to have a competition for truth because you also said like what what is true it's because what may be true to some may not be viewed as true to others

6:55but you want to have the closest approximation of that so I think the community notes thing is very powerful I think we are trying to have as many organizations and people and institutions verified as being legitimately those people and organizations is is important and to have the organizational affiliation clearly identified so that if you want to find out if somebody's actually if an account is actually saved from A Member of Parliament or a journalist or if let's say if a

7:30Twitter handle is actually belongs to say Disney Corporation or something like that you you can you can go on Twitter and it's it's sort of an identity layer of the internet and you you can confirm that that is a fact the case and and I think once you've got these sort of interlocking sort of identities it's actually very hard to be deceptive in that case because and it's also you have a reputation to protect at that point so I think then people are far more likely to be measured in their response and we will be more reasonable since they have reputational value at that point

8:11so these are some of the ideas that I have and you know I'm not saying that for sure it will succeed or that it's going to be perfect but I I I I am confident that it will over time you know head into in a good direction and I think that the evidence for that will be do people find it useful you know as we're measuring sort of the you know total user minutes but not just user minutes unregretted use minutes which I think that the key figure of marriage you know because well for example Tick Tock has a very high usage I often hear people say well I spent two hours on Tick Tock but I regret those two hours I'm not just trying to knock Tick Tock but it's just we don't want that to

9:00be the case with Twitter we want to say like okay you spent half an hour on Twitter that you found it to be useful and entertaining and a good thing in your life and ultimately be a force for good for civilization that that's the aspiration thank you Elon we have over 150 government within the world government Summit global leaders they have 8 billion customer their citizen how government can use Twitter bitter to

9:36service citizens yeah well I I think generally I would recommend you know really communicating a lot on Twitter and I I think it's good for people to speak in in their voice as opposed to how they think they should speak you know like you know sometimes like people think well I I should speak in this like

10:10but it ends up sounding I think somewhat at times somewhat stiff but and not not real you know like if you read a press release from Corporation it just sounds like propaganda so I would encourage CEOs and of companies and you know legislators and you know ministers and so forth to to speak authentically and to you

10:37know if there's a say a particular policy to explain it and and I I think there's you know sometimes a concern about criticism but I I think at the end of the day you know it's having some some criticism is is fine you know it's not that it's really not that bad I mean I I I'm constantly attacked on on Twitter frankly and I I don't mind it's it's you know you have to be somewhat thick skinned I suppose at times you know just because you're gonna they're really trying to twist a knife but but I but I think I think just like I said just as a form for communication

11:20it's great and and I would just encourage more communication and and like I said to to sort of speak in an authentic voice like like sometimes people will have someone else be their sort of Twitter manager or something like that and I think I it's just people should just do their own tweets you know it sounds it and and like sometimes you make a mistake or something it's fine but I think just doing your own tweets just like you would do your own you know you give a talk here or you would you know have a meeting at a summit or something I think that's that's the way to do it is is to actually do do the tweets yourself and and convey the message that you want

12:09directly yeah so but you know I mean one thing I should say I know this is called the world government Summit but I think we should be maybe a little bit concerned about actually becoming too much of a single World Government if I may say that we want to avoid creating a civilizational risk by having frankly this may sound a little odd too much cooperation between governments you know if you know if you look at say the history and the rise and fall of civilizations the really all throughout history civilizations have risen and Fallen but it hasn't meant the Doom of humanity as a whole because they've been they've been all these separate civilizations

12:58that were separated by great distances and so you know say like while Rome was falling you know Islam was rising and so you had like a you know the the sort of caliphate doing incredibly well while Rome was doing terribly and that actually ended up being a source of preservation of knowledge and and many as scientific advancements and so so I think we want to be a little bit cautious about being too much of a will of a single civilization because if we are too much of a single civilization then if we're if the whole the whole thing May collapse I don't know obviously not suggesting war or anything like that but I think we want to be a little bit wary of actually

13:50cooperating too much it sounds a little odd but but we just we want to have some amount of civilizational diversity such that if if something does go wrong with some part of civilization that the whole thing doesn't collapse and and you know Humanity keeps moving forward thank you I you see I I hear you I

14:13agree and disagree with you a certain point and I think you know it's today people they don't fight with sword anymore I mean they have nuclear weapons so there is there is this conflict the whole civilization will be gone the whole human civilization will be gone and what we are trying to do here at the Emirates actually is to do exactly what you are saying we have 180 nationality you have every single race every single religion and we are trying to create a model that show the world that doesn't matter who you are what your color what's your religion where you're from you Humanity can live in peace and harmony yeah I mean that would be good yeah my my

15:01last question I'll go to Twitter again then we'll move out of Twitter if you allowed us I mean you've been running Twitter as at the chairman at the owner at the CEO and that's take a lot of time did you identify a CEO and when you are going to hire him well I I think I need to stabilize the organization and just make sure it's in a financially healthy place and that the the product roadmap is clearly laid out so I don't know I'm guessing probably towards the end of this year which would be good timing to find someone else to run the company because I think it should be in a stable position around you know at the end of this year Elon if we move to other subjects I mean at the

15:56summit here we have speaker who speak about state of the word state of geopolitical of the world for the next 10 decade state of the economy of the world you know now and in the next 10 years if I ask you about the state of Technology if you can elaborate a bit then brief us how do you see technology in the next 10 years from now

16:26let's see 10 years it's always difficult to predict technology with Precision especially over a 10-year time frame when it is changing so much I mean there's there's obviously the transition to sustainable energy with the solar wind batteries and electric vehicles and and that that's if you look at the percentage growth of that that is a very high percentage growth although because of the massive industrial base of of the current sort of fossil fuel

17:02economy it it even like even if all for example if electric cars were 100 of production immediately it would take 20 years to replace the fleet so this is still something that is quite gradual you know it's measured in just a few you know 30 30 40 years type of time frame on a more sort of near-term time frame I think artificial intelligence is something we need to be quite concerned about and really be

17:34attentive to the safety of of AI you mentioned a chat gbt earlier you know I I played a significant role in the creation of openai essentially at the time I was concerned that Google was not paying enough attention to AI safety and so I I with a number of other people created opening and although initially it was created as an open source non-profit and now it is closed source and for-profit I don't have any stake in open AI anymore nor am I on the board nor do I control it in any way but the chat GPS here I think has Illustrated to people just how advanced AI has become

18:28because the AI has been Advanced for a while it just didn't have a user interface that was accessible to most people so what really chat gbt has done is just put an accessible user interface on AI technology that is this has been present for a few years and there are much more advanced versions for that that are coming out so I think we you know I think we need to really be I think we need to regulate AIC quite frankly because if you think of any technology which is potentially a risk to civil to to people like if it's an aircraft or your cars or medicine we have regulatory bodies that oversee the public safety of cars and planes and Medicine and

19:30I think we should probably we should have a a similar sort of regulatory oversight for artificial intelligence because it is I think actually a bigger risk to society than cars or planes or medicine so and this basically have to slow down AI a little bit but I think that that might also be a good thing the the challenge here is that a government regulatory authorities tend to be set up in reaction to something bad that happened so if you look at say aircraft or or cars you know the cars were unregulated at the beginning aircraft were unregulated but they had lots of you know airplane crashes and in some cases manufacturers that were cutting Corners

20:24and and a lot of people were dying so they the public was not happy about that and so they established a regulatory authority to improve safety and now commercial airliners are extremely safe if you were to drive somewhere it's a safety for a mile of a commercial airliner is better than a car and cars are also extremely safe compared to where they used to be so but if you say if you look at say the introduction of seat belts the Auto industry fought the introduction of seat belts as a safety measure for I think 10 or 15 years before finally The Regulators made them put seat belts and cars and that greatly improved the safety of cars and that airbags were another big

21:18Improvement in safety so my concern is that with AI if if there's something rather if something goes wrong the reaction might be too slow from a regulatory standpoint so I I would say like it you know if it's like one of the biggest risks to the future of civilization positive or negative it has great great promise great capability but it also with that comes great danger I mean like I say nuclear it you know just discovery of sort of nuclear physics you had nuclear power generation but also nuclear bombs so anyway I think we should be quite concerned about it and we should have some regulation of what is it if I fundamentally a risk to the public

22:15yeah very great let me move to another subject Elon education I mean you have your own philosophy about education with AI education might change dramatically when you tell us briefly about your philosophy of education and number two do any 12 years of schooling and four years of University

22:44well with respect to education I think in general some things that we could do to make it more compelling would be to explain to Children why we are teaching a particular subject so the human mind has evoked to really forget anything that it deems unimportant so if that human memory is really quite quite bad relative to the memory of your phone your phone is can remember the entire contents of an encyclopedia down to last a letter in pixel but human memories is terrible by comparison so the mind is constantly trying to forget things actually so if you but if you explain the the why why a subject is being taught that will then establish relevance

23:38and it's much more likely to result in motivation for kids and and then also if you if you teach knowledge especially in the Sciences as solutions to a problem it's much more effective so like let's say you're trying to understand an internal combustion engine well it's actually better to sort of take that apart and and then say okay well what tools do we need to use to take it apart we need a wrench and screwdriver and various other things to take it apart well then then you understand that the reason for the tools and so like for for mathematics and and it's I like tools and in physics and Engineering but if you if you but if you teach to the problem and say and then you

24:28understand then you establish the relevance of the tools then you it's actually much easier to remember mathematics and physics because they'll help explain how the world works as opposed to teaching them without explaining why and simply teaching them it's like instead of having teaching to the problem teaching currently people teach the tool it would be like having a course on on screwdrivers or course on wrenches but not understanding why you have a course you're learning about screwdrivers and wrenches I think this is really quite a fundamental principle that should be applied in education and and I think sometimes we do we do teach classes that are that children do

25:13not find useful and and where the answer to the why is actually not going to be a very good answer you know most people I think will do not find advanced mathematics useful and are unlikely to find it useful in their life or the elements that they do find useful could be taught very quickly as general principles I also think that critical thinking is something that should be taught to children out of a relatively young age as effectively like a mental firewall

25:48to really think about when somebody tells you something is it cogent is it true or what is the probability that it is true and so that you can to be taught to reject things that are untrue or more likely to be untrue and favor things that are more likely to be true critical thinking I think is very helpful for people to learn so is it 12 years of schooling you are with or without 12 fears

26:28well yours is a long time I suppose I mean humans just do take a long time to mature so there's emotional maturity physical maturity and mental maturity that is happening simultaneously with the education I suppose it could be done in 10 years perhaps it does not need 12. but but then is someone mature at age 16 they're more likely to be mature at age 18. so I guess 12 years is probably not bad we probably don't need an additional four or five or six years in in college

27:12and university that seems probably excessive I think we'll probably shave a few years off and be fine okay kids will love it by the way yeah but you know social media but we spend so many hours on social media I mean the average sometime in certain country three four hours in social media and sometimes when we go to our kids we we see them spending also long hours part of his knowledge do you have any rule for your kids I mean how much they can spend in social media

27:57right you know I'm certainly not try to restrict social media for my kids although that might have been a mistake depending on which kid it is that I mean they've really been programmed by Reddit and YouTube I'd say more than anything else Reddit in YouTube I think probably I would limit social media a bit more than I have in the past

28:29and just you know let's take note of what they're watching because I think you know at this point they're being like programmed by some social media algorithm which you may or may not agree with so I I think probably one is to supervise children's use of social media and be wary of them getting programmed by some algorithm written in the Silicon Valley you know which you know maybe not be what you want Elon you've been working very hard I mean since six years ago we met you you look much younger by the way about six years ago thanks but I I know that he's been working for almost 20 hours a day you sleep and the sofa in the office maybe Twitter office Tesla office you

29:23told me once we I I was with you at Tesla's office how do you balance your life I mean with this stress with so many different you know you're on so many different companies how do you balance it well I I mean I I should point out that I a 20-hour workday is relatively unusual and and rather painful but but I I do sleep six hours a night so and if I sleep less than six hours and I I find that I am I might be awake longer but I get less done so

29:57but I I do have a worker ridiculous amount I think relative to most people and in that it's pretty much seven days a week and mostly from when I wake up to when I go to sleep I'm not suggesting this is good for everyone and I think frankly I would like to work a bit less on that so

30:21say Tesla Tesla went through some very difficult times where it was on the Ragged edge of survival and and really if I if I didn't give it everything I got it the company could have easily gone bankrupt it was really on the verge of bankruptcy for quite a while I don't mean to suggest complacency at this point but you know it does require much less work to operate Tesla Now versus say in the 2017 to 2019 time frame

30:54and it's not a mortal risk of survives it's achieved economies of scale that make it you know not on the Ragged edge of survival and then the SpaceX also has the strong team and is able to make a lot of progress even if I spent less time there it does help if I spend time there but you know it keeps making progress even if I don't Twitter is still somewhat of a startup in reverse and so there's a lot of work required here to get Twitter to a sort of a stable position

31:33and like I said to really build the engine of engineering of software engineering at a Twitter and have it you know really have it likes to do a great product roadmap and and the people in place to implement that product roadmap and so it is not my intention to work like crazy you know I mean I think I still I I got comfortable with a mere 80 hour work week would be fine all right

32:07is what I would aspire to thank you we are running out of time I have one last question I have to ask you sure three UFO B shot one over Alaska Lake Huron and Canada alien no alien I don't think it's aliens no okay I mean I do find the whole question of of aliens a very interesting one you know what is typically called the Fermi Paradox which is that if the universe is really as old as it as a science seems to think it is and as and the the where are the aliens have really been around for 13.8 billion years if so where should there be aliens all over the place why do the crazy thing is I've seen no evidence of of alien technology or any alien life whatsoever

33:08and I think I I think I know you know SpaceX we do a lot I mean I think I know I don't think anyone knows more about space you know than than me or at least the space technology so but I think it's actually a troubling thing if there are no aliens as well which is that all that what what that actually could mean then is that sort

33:36of civilization and Consciousness is like a tiny candle in a vast darkness and and a very vulnerable tiny candle that could easily get blown out and I think we should therefore take great care with what may very well be this tiny handle in a vast darkness and make sure that it does not go out and that we extend the light of Consciousness beyond Earth and do everything we can to ensure that the light of Consciousness does not go out Elon we run out of time thank you very much and hopefully to see you next year with us here in the Emirates with your family sounds good and thanks so much thanks again for having me thank you thank you we'll see you bye-bye

资源与出处

相关视频

同频道、同主题或相邻年份