推荐微信扫码登录以获得更好的体验!另:🎉建议更新到最新版,以体验新功能!🎉 版本更新记录→

面对全球监管风暴,周受资全英文重磅专访:TikTok 的破局

字幕摘录

时间英文中文
0:00Okay, good morning everybody and show thanks for taking the time to be with us好了,早安,大家好,谢谢你们抽空陪我们
0:05before we在我们之前
0:06Go into the chat.进入聊天。
0:08I thought I'd like to have the audience我想请观众
0:11Give some of their views.给他们一些看法。
0:14So there is a question for the poll.民意调查有个问题
0:16Can we bring the poll question up, please?我们能把民意测验的问题提出来吗?
0:28Maybe are you all gonna bring the poll question up?也许你们都会提出民意测验的问题?
0:40Okay, if not, I tell you what I'm gonna do.好吧,如果没有, 我告诉你我该怎么办。
0:42All right好吧,我们走
展开字幕全文(456 条)
序号英文中文
1Okay, good morning everybody and show thanks for taking the time to be with us好了,早安,大家好,谢谢你们抽空陪我们
2before we在我们之前
3Go into the chat.进入聊天。
4I thought I'd like to have the audience我想请观众
5Give some of their views.给他们一些看法。
6So there is a question for the poll.民意调查有个问题
7Can we bring the poll question up, please?我们能把民意测验的问题提出来吗?
8Maybe are you all gonna bring the poll question up?也许你们都会提出民意测验的问题?
9Okay, if not, I tell you what I'm gonna do.好吧,如果没有, 我告诉你我该怎么办。
10All right好吧,我们走
11Just the ah就是那个啊
12Okay还好
13Bring it up.举起来.
14Okay.摆
15Good.不错
16So we we asked this question earlier on of many in the audience and we asked in the age of AI我们之前问过很多观众的问题, 我们在AI时代问过
17What will matter more and not less and so there were a few questions or a few answers还有什么比这更重要的,不是更少的, 所以有一些问题或几个答案
18We had and we can still see some answers earlier on human creativity我们以前有过,现在还能看到一些关于人类创造力的答案
19human judgment人类的判断
20human curiosity人类好奇心
21human relationships or human purpose人际关系或人类目的
22So for those of you who have not所以你们中没有
23Put in your answers and you still would like to put them in you can把你的答案,你仍然想 把它们,你可以
24Try and scan and put your answers in now.试着扫描一下,把答案放进去
25I think some of you had done it a bit earlier我想你们中有些人早点做了
26Okay, let's give it a still moving still moving.好吧,让我们让它静静地移动。
27Let me be respectful of those who are scanning and putting your answers in让我尊重那些 扫描和把你的答案
28Okay还好
29Let's give it another再来一次
30Few seconds or so.几秒钟左右。
31Oh瓳
32Still moving.还动.
33Okay.摆
34All right.好吧,我们走。
35I think from here we can see from these我想从这里我们可以看到这些
36Poll results, so thank you very much for your participation looks like a human judgment投票结果,所以非常感谢你的参与 看起来像是人类的判断
37Has got a thirty six point nine percent百分之三十六 百分之九
38Now thirty six point six.现在三十六点六分
39So it's between human judgment所以这是在人类的判断之间
40human relationships and人际关系和
41Then kind of human creativity and human purpose kind of a on the edge然后一种人类的创造力 和人类的目的 一种边缘
42So thank you very much for your participation now非常感谢你的参与
43I've given all of you one answer to give so might have been a bit difficult我给大家一个答案 可能有点难
44But for sure since it's our special guest不过既然是我们特别的客人
45I thought I'll give him a little bit more of a K long我想我会给他多一点K长
46You know Singapore terms a bit of a long thing, right?你知道新加坡说得有点长吧?
47So sure you can have more than one answer确定你能有一个以上的答案
48But what do you think about this?但你觉得怎么样?
49I think it's such an interesting我觉得这很有趣
50Survey and first of all, thank you DBS for having me.首先 感谢DBS邀请我
51Thank you Sushant for the introduction谢谢你的介绍
52I am from Singapore and I'm quite我来自新加坡,我相当
53Transparency I am too born bread through blue透明,我太天生的面包通过蓝色
54I think I think what is really interesting about this survey is我觉得这次调查真正有趣的是
55We don't have to question the definition of human here, right?我们不必质疑人类的定义,对吧?
56I think human is a very obviously a very broad term and between human to human even all of us in this room我认为人类显然是一个非常宽泛的术语,在这个房间里的我们所有人都是人与人之间。
57I think it is fair to say our judgment is different我认为说我们的判断是不同的是公平的
58We're gonna have a varied judgment between myself yourself, you know people in various corners of the room我们对自己会有不同的判断, 你知道人们在房间里的各个角落
59so the fact that we所以,事实上我们
60able to能够
61You know very quickly come to a consensus that there's such a thing called human judgment human, you know relationships human creativity你知道很快就达成了一个共识 一种叫做人类判断的东西 你知道人类的创造力
62I think it juxtaposes with what we are dealing with which is something that's not我认为它与我们所处理的事情是并存的
63Innately biologically like us and may behave in a way that's very different from us.天生就和我们一样 行为方式可能跟我们完全不同
64I一、导 言
65I do agree with basically, you know, all of the above我基本上同意,你知道的,所有以上
66I think all of this these five things will matter a lot.我想这五件事都很重要
67I think for judgment, you know, there's a lot of talk about我想为了判断,你知道, 有很多谈论
68AI alignment at this moment in time and of course, there's some very extreme examples that have been given out thereAI在此时此刻的校对,当然, 有一些非常极端的例子 在那里,
69I think there's a paperclip example.我想这里有个纸屑的例子
70That's very famous这很出名
71And I may be butchering this analogy, but it's something along the lines of you ask the machines to produce more paperclips and it ends up我也许在宰了这个类比, 但这是你这样想的, 要求机器生产更多的纸片,结果
72Doing that only one thing which is to completely overweight on production of paperclips at the expense of everything else只做一件事,就是完全超重 生产纸片,牺牲一切
73so that这样
74You know, I think we are grappling a lot with how do we?你知道,我觉得我们有很多 与我们如何努力?
75As we're dealing with something that's so powerful as Sushant alluded to things that can solve very complex mathematical problems当我们面对着苏珊特所说的 能够解决非常复杂的数学问题的力量
76Now I was chatting with some other entrepreneurs over the weekend.周末我和其他一些企业家聊天
77My view is that in the next two to three years我的看法是,在未来的两三年里
78We're gonna have a real scientific breakthrough我们要有一个真正的科学突破
79physics biotech物理学生物技术
80Chemistry material science because of AI it's very very very possible化学材料科学 因为AI,这是非常可能的
81And if that happens, then you know, the question would be how do you align?如果那样的话,你就会知道, 问题在于你如何配合?
82The models, how do you align the machines to be doing things that?模型,你如何调整机器 做那些事情?
83It's obvious to us as human beings like this is very obvious that these are the boundaries of what you should do morally and ethically作为这样的人,我们很显然 这些是你们应该做的道德和伦理的界限
84And these are the boundaries of what you should not do.这些都是你不应该做的事的界限。
85I don't think we have an answer to be honest at this moment in time此时此刻,我想我们没有老实回答
86I think it really is a very complex issue我认为这是个非常复杂的问题
87But it will it will matter in the age of AI.但它会在AI时代有所影响.
88So so that's on judgment on creativity所以这是对创造力的判断
89For a tick tock it is very relevant对于一个滴答,它非常相关
90Just as way of introduction for those who don't know tick tock is a short video app对不知道滴答的人的介绍 只是一个短视频应用
91If you don't have it, there's Wi-Fi you can download it is free如果你没有,有Wi-Fi 你可以免费下载
92to use用于
93Apart from tick tock, you know at bike dance.除了滴答,你知道在自行车舞。
94We also produce our own AI models我们还生产自己的AI模型
95Right now for our video generation model is called seed dance is现在,我们的视频生成模型 叫做种子舞
96Probably one of the better ones today.也许今天是更好的之一。
97Our industry term is SOTA.我们的行业术语是SOTA。
98It's state of the art and这是最新艺术和
99Our large language model is我们的大语言模型是
100Not that it's pretty good.不是说它相当不错。
101It's pretty up there and we have built those models into two businesses one上面相当漂亮,我们把这些模型建成两个企业
102Which is the power our internal consumer apps like take out like doing like dope out in China and the other is to sell it as an API我们内部的消费者应用软件就像在中国做毒品一样 拥有的能量是作为API出售
103So through our AI team, so if anyone's interested, you know, of course studying sushi and I'm going to use this opportunity to do some advertising for所以,通过我们的AI团队, 所以如果任何人有兴趣, 你知道,当然,学习寿司 我要利用这个机会做一些广告
104model as a service business作为服务业务的模式
105But for creativity itself, it's it's quite interesting because you know with但对于创造力本身来说 这很有趣 因为你知道
106We have allowed our users to produce a lot of very realistic AI videos using our models我们让用户用我们的模型 制作许多非常现实的AI视频
107I have produced a few myself.我亲手制作了几本
108We have published it我们已出版
109There is still a an imbalance between the percentage of videos that are being created正在创作的视频比例之间仍然不平衡。
110they are AI generated and the number of views and the percentage of views that are being seen and它们是大赦国际生成的,意见的数量和所看到意见的百分比,以及
111It's lopsided in the way that there are more people as a percentage creating AI content than there are views of it它的偏差在于 创造AI内容的人比例比有观点要多
112So I think there's still a very strong所以我觉得还有很强的力量
113Desire because the views is powered also by AI by a recommendation algorithm and it's a very democratic experience渴望因为观点也是由AI通过推荐算法赋予力量的,而这是一种非常民主的经验.
114So it really shows people what they want to see所以它真的向人们展示了他们想看到的
115So what we took away from this is I think there's a lot of所以,我们从这里拿走的是 我认为有很多
116Experiment with using AI to create more more content and a lot of it is a lot more creative than you can ever imagine用人工智能创造更多内容的实验 并且很多都是你想象不到的
117But at the same time, I think human adoption of just watching AI content is not the same at this moment in time但与此同时,我觉得人类通过 只是看AI的内容是不一样的 在这个时候
118It's still low single digit percentage in terms of views.其收视率仍低于单位数百分比.
119So I think people still value a lot of authenticity所以我认为人们仍然很重视真实性
120When I see a video and you know, it's very obviously not当我看到一个视频 你知道, 它非常明显不是
121Real it's very obviously artificial like sort of artificially generated真实的,它显然是人工的 像某种人工生成
122I think there's still a little bit of pushback in terms of consumption people still want to see very real very authentic things我觉得在消费方面还有一点退缩 人们仍然想看到非常真实的东西
123So I agree that human creativity will continue to be important.因此,我同意人类的创造力仍然很重要。
124I'll pause here.我会在这里暂停。
125Yeah, thank you show是啊,谢谢你显示
126You've been very generous with commenting on that你一直很慷慨的评论
127So as we as we have heard from what show has said actually all of these various characteristics here正如我们从节目中听到的 节目中讲到的,这些不同的特点
128Which have got some people voting for are indeed certain aspects让一些人投票支持的 确实是某些方面
129That we human beings I think can can and will have to work我们人类,我认为可以 并且将不得不工作
130With the evolution of this technology随着这种技术的发展
131So in fact actually show has answered one of the questions I was gonna ask him about所以,实际上显示已经回答了一个问题 我本来要问他关于
132Whether in this world where AI is increasingly being used to recommend to influence and to make decisions在世界中,AI是否越来越多地被用来建议影响和作出决定?
133What is meaningfully human?什么是有意义的人类?
134Or what remains meaningfully human?或者什么是有意义的人类?
135I think sure has given quite a good我想这给人很大好处
136Kind of explanation to this but if I may just build on that a little bit a little bit more有点解释,但请允许我再多一点
137Therefore in light of this given that we're at this stage of AI evolution因此,考虑到我们正处于AI进化阶段
138Where as you've mentioned, you know things are still evolving你提到的地方,你知道事情还在演变
139Authenticity is still very important in content creation认证在内容创建中仍然非常重要
140What therefore should human beings actually work on in?因此,人类应实际从事哪些工作?
141Developing our capabilities in order to remain successful in this age of AI.发展我们的能力,以便在这个AI时代保持成功.
142I think there is increasingly我认为有越来越多的
143the belief that认为
144This should be a very it should be a human AI sort of应该是人类的人工智能
145Combination rather than letting the machines take complete full reign of all decisions联合起来,而不是让机器完全掌握一切决定
146No, of course increasingly with as technology gets better不,当然随着技术的提高, 技术的不断进步
147I think and as we build the infrastructure to let the agents do more things for us and we build their trust and the guardrails我认为,当我们建设基础设施,让特工们为我们做更多的事情时,我们建立他们的信任和护栏。
148Probably what will happen in the long term is we will allow the agents to do more and more things autonomously可能从长远来看 我们会让特工们 自主地做更多事情
149I think this is probably the trend but in the short term, I think a short meaning at least for the next我认为这可能是趋势,但在短期内,我认为一个简短的含义至少对于下一个
150Two three years.两年三年.
151I think it's going to matter a lot我觉得这很重要
152The skills that you develop to be able to use AI very effectively.你为了能够非常有效地使用AI而发展的技能.
153I think that's probably the most important thing我想这可能是最重要的事情
154for us internally like对我们内部一样
155What we are encouraging not just our engineers我们所鼓励的不只是我们的工程师
156But also, you know our lawyers and people working in corporate functions do is to have more first-hand experience并且,你知道我们的律师 和在公司工作的人 是拥有更多的第一手经验
157Because if you don't use these and it's changing so quickly if you don't use these models and you don't use these products yourself因为如果你不使用这些 并且它正在变化如此之快 如果你不使用这些模型
158You cannot figure out what exactly it can or cannot do你无法知道它到底能做什么或不能做什么
159just by reading about it, so I think one of the只是通过阅读它, 所以我认为其中之一
160Key characteristics, you know, at least for the next couple of years.关键特征,你知道, 至少在未来几年。
161There will be really important is这很重要
162Being hands-on by being curious being very hands-on, you know open claw came out以好奇为己任 你知道打开的爪子出来了
163You know, we encourage people to try it obviously understand the boundaries of you know, the security aspects of it你知道,我们鼓励人们尝试它 明显理解你的界限,它的安全方面
164But try it for yourself.但自试下.
165So you understand what the hype is about you understand所以,你明白 胡言乱语是什么 你明白
166You know where the boundaries of this technology can be and it's not some你知道这种技术的界限在哪里 而不是某些
167Mythical thing that can do everything at this moment in time is not此时此刻能做一切的神话不是
168So I think a lot of that所以我觉得很多
169at least for internal employees至少对内部雇员而言
170We have developed a lot of programs to allow them to have access to frontier models.我们制定了许多方案,使他们能够接触前沿模式。
171We give them some我们给他们一些
172Programs that we pay for some of their token consumption我们支付他们一些象征性消费的方案
173not only for work use but also a little bit of personal consumption as well and the encouragement is to make sure people are不仅为了工作 也为了一点个人消费 并且鼓励人们
174Being very hands-on and really using these非常亲切和真正使用这些
175New tools as they come into the market to stay on top of it新的工具,当他们进入市场 留在上面
176I think that's probably the most important thing right now.我认为这可能是最重要的 现在的事情。
177Thank you.谢谢
178Thank you谢谢
179I I can't agree more with many of these things you're saying right about still having我完全同意你说的这些话
180Really humans at a center making the making the judgment deciding what to do what not to do真正的人类 在一个中心 做出判断 决定做什么 不做什么
181And in this case about really embracing AI而在这种情况下,关于真正的拥抱AI
182I think sometimes in life people take a very extreme stand.我觉得生活中有时人们会采取非常极端的立场.
183It's either humans or machines是人还是机器
184Sometimes people take a therefore even more extreme stand to totally reject it有时人们会采取更极端的立场 完全拒绝
185protest against it and some抗议和反对
186Overly hyped up with the embracing of it.超乎寻常地被拥抱着
187So I think that ability for us to balance所以我认为我们有能力平衡
188I agree with you is actually very very important.我同意你的观点,其实非常重要
189Yeah on that对啊
190I I think it was some I can't remember some somebody else said this I think short term我想这是一些我不记得了 别人说,我认为短期
191We are probably overestimating overestimating it but long term with that probably underestimating and the trend is very very clear我们可能高估了它,但从长远来看,可能低估了它,而且趋势非常明显。
192it's这是
193The we are only at the beginning of what you know artificial intelligence can probably do and我们才刚刚开始 你知道人工智能可能能做的
194If you look at you know, where the trend is the progress is particularly in areas of science如果你看看,你知道, 哪里的趋势是 进步特别在科学领域
195It's really trending on a very steep curve at this moment in time此时此刻,它正向一个非常陡峭的曲线发展
196So I think long term, you know, we probably don't fully understand the profound impact所以我想从长远来看,我们可能不完全理解 深刻的影响
197This is going to have across our human lives, you know over the long term by the short term.这将会贯穿我们的生命, 从长远来看,从短期来看。
198It's you know这是你知道的
199Human beings need time to adopt these kind of technology.人类需要时间来采用这种技术。
200We talked about AI coding in in our industry我们谈了我们行业的AI编码
201We are probably one of the first to get disrupted by AI, especially in terms of coding itself我们可能是第一个被AI打乱的人之一, 特别是在编码方面
202But coding really only takes up 30% of most of the engineers time a lot of the other stuff但编码只占用30%的工程师时间 其他很多东西
203It's still very human, you know writing your engineering plan planning, you know talking to a product manager and a lot of their coordination它仍然非常人性化,你知道写你的工程计划, 你知道与一个产品经理交谈, 和他们的很多协调
204Yes, you can see how as AI tools get better that also goes away是的,你可以看到 AI 工具如何变得更好 也走了
205But it will probably take longer than people expect但可能比人们预期的要长
206You know, there's still a lot of you want to let you know, you want to double-check a little bit你知道,还有很多人想告诉你, 你想检查一下
207You're not going to deploy your code without checking at all.你不检查一下就不会部署密码
208That's that's really not you know, advisable这真的不是你知道的,可取的
209So short term is probably this kind of you know sense in the gear所以,短期可能是 这种你 知道感在齿轮
210It's probably going to slow down progress可能会放慢进度
211But I think it's risky to think that just because it's slow in a shorter term但我觉得这样认为很冒险 只是因为很慢
212That it will not become very important in the long term.从长远来看,它不会变得非常重要。
213It's also that's the wrong conclusion to take away这也是错误的结论 带走
214I agree and that's why it's very important for us to actually start embracing building it because at some point we believe that there will我同意 这就是为什么我们真正开始拥抱它是非常重要的 因为有时候我们相信
215Be an inflection and when the inflection comes it's very important for us to be ready做个坏事,当坏事降临时, 对我们来说,做好准备是非常重要的
216personally as well as in the businesses that we run if I may then just move on a little bit more because无论是个人还是我们经营的生意... ...如果我能再继续前进一点... ...因为...
217Actually, as you mentioned, you know, it's it's probably something which human beings would take a bit more time to learn and adopt and therefore事实上,正如你提到的,你知道, 它可能是一些东西, 人类会花更多的时间学习和收养, 因此,
218From what you've seen not just in the business you run but in your interactions with people从你所看到的,不只是你经营的生意, 而是你与人的互动
219Have you where have you really seen AI?你真的在哪里见过AI?
220Enable just ordinary people to achieve things that they previously never thought was possible or many many aspects让普通人能够实现他们以前从未想过的可能或许多方面
221internally, we try to simplify this into a内部,我们试图 简化这一点为
2222x2 matrix2x2 矩阵
223You know if you take things that are你知道如果你拿的东西
224Routine and very non-routine, you know常规和非常非常规,你知道的
225Very sort of differentiated as one axis and you take things that kind of high stakes and low stakes as the other axis有点像一个轴,你把 那种高桩和低桩作为另一个轴
226You very quickly find that things that are routine and low stakes really honestly the agents can automate and do very very quickly你很快就发现,那些例行公事和低赌注 老实说,代理可以自动化 并非常迅速地做
227Anything that's high higher stakes.任何高赌注的东西
228You probably want to be a bit more careful你也许更想小心一点
229But the frontier models can probably do most of that too anything that's very high stakes and very very但边疆模型可能也能做 大部分那些非常重要和非常
230Differentiated a very high touch.差别甚高触.
231I think those are the things that you probably need still need a lot of human involvement我认为这些是你可能需要的 还需要很多人类参与
232Let me give you one example让我举个例子
233One thing about tiktok is we have a lot of users around the world todayTiktok的一件事就是 我们今天在世界各地有很多用户
234Sushant mentioned you safety, you know, it's obviously a very big thing for us now in order for us to understand say a 14 year old苏珊提到你的安全,你知道, 这显然是一个非常重大的事情 现在,为了让我们明白 说一个14岁
235experience in Nigeria, you know在尼日利亚的经历,你知道的
236The French speaking experience in Nigeria for a 14 year old, you know in the past14岁在尼日利亚的法语经验,你知道的
237Yes, you know we can have our trust and safety human beings go set up an account, you know test the account是的,你知道我们可以有我们的信任和安全的人 去建立一个账户,你知道测试账户
238Watch a hundred videos write a report about it看一百个视频写报告
239Now I can just deploy a thousand agents and do it overnight and the next morning it will just generate its own report现在,我可以部署一千名特工 并做一夜, 第二天早上 它会只是生成自己的报告
240What is the content you saw, you know, whether it's how much of it is violated against our guidelines你看到的内容是什么,你知道, 它是不是有多少违反我们的准则
241I mean this whole thing is automated when we talk about content moderation我的意思是,这整件事是自动化的 当我们谈论内容节制
242we have a very我们有一个非常
243significant live streaming platform and life by definition is quite difficult to to moderate because it's life and there are从定义上说,重要的直播平台和生命很难温和,因为它是生命和生命
244Tens of millions of live streams that's happening at any moment in time in the past过去任何时刻都会有数千万条活溪流
245You know without a lot of the new AI tools in the market你知道市场上没有很多新的AI工具
246It is quite difficult you do some random random sampling for live streaming rooms做一些随机的随机采样 活流室是很困难的
247But it's quite difficult to be very very precise and very accurate at the same time但很难做到非常精确和准确
248but right now with this video recognize但是现在这段录像能认出来
249VLM models you are able to just very quickly take a live streaming videoVLM模型,你能够快速地拍下视频
250Convert it into a text narrative and then match that against your community guidelines转换成文字说明,然后与您的社区指南相匹配
251Things like, you know, somebody's driving a car比如说,有人在开车
252The AI is able to see that the speedometer is tracking the car at 160 kilometers per hourAI可以看到速度计以每小时160公里的速度追踪这辆车
253A human being might miss that, you know, if it's a hundred sixty kilometers per hour is dangerous driving人可能会错过,你知道, 如果它是每小时一百六十公里是危险的驾驶
254You know, it's things like that that AI has tremendously improved our capability to do a lot of our work你知道,事情是这样的,AI 已经极大地提高了我们的能力 做我们的许多工作
255So I mean I can give you a lot more examples like that.所以我可以给你更多的例子。
256So on the one hand, you know, we所以一方面,你知道,我们
257You know你知道吗
258Obviously like everybody else are mindful of the new risks that it could bring to the market the safeguards we need很明显,像其他人一样, 也意识到了新的风险 它可以带给市场 我们需要的保障
259In how do we understand the cyber security risks it brings on the other hand, we are tremendously excited by how much我们如何理解网络安全带来的风险,
260Productivity it enables within the company on a personal note公司内部的生产力 以个人名义
261How has AI actually helped you already kind of amplify your own?AI怎么帮助你 已经有点扩大你自己?
262capabilities own creativity I自我创造力
263I use AI as a lot as a sounding board我经常用人工智能做音响板
264For example, you know recently became a US taxpayer and the US tax, you know regime is practically impossible to understand例如,你知道最近 成为美国纳税人和美国的税, 你知道政权几乎不可能理解
265You know without the help of AI.你知道,没有AI的帮助。
266So with AI has really helped me a lot所以用人工智能帮了我很多忙
267I've also become as become my personal nutritionist我也变成了我的营养师
268I take a picture of you know food that I eat and it tells me, you know我拍了张照片,你知道我吃的食物 它告诉我,你知道
269Don't eat so much of this eat more of that别吃这么多,多吃点
270It just feels like it's a very nice companion to have on a personal level like a super smart这感觉就像一个非常好的伴侣 在个人层面 像超级聪明
271Person whom you can always have a conversation with and and on the health side, for example, I am比如说,我...
272I broke my shoulder earlier this year and I maybe had 200 questions for my surgeon今年早些时候我肩膀断了,我也许有200个问题要问外科医生
273But AI answered 90% of it.但AI回答其中的90%.
274So I saved I saved the surgeon 90% of a questioning time所以我救了医生90%的审讯时间
275So I use cases like this.所以我用这样的案子。
276I think it's just tremendously helpful我觉得这很有帮助
277Yeah, actually I just met show yesterday at the paddock right and and to me是的,其实我昨天刚在仓库见了个面
278My impression of show every time I see he looks pretty much very fit like that我每次看到他都觉得他很合身
279And I hadn't realized he's lost some ways gotten a lot fitter and say how he said AI and exactly what he just shared我还没意识到 他失去了一些方法 得到了很多适切 并说他怎么说AI 和他刚刚分享的东西
280About how some sub how his agent will stop him from eating what he should have stopped eating.他的经纪人会如何阻止他吃他应该吃的东西
281It will remind me它会提醒我
282It will remind me.它会提醒我。
283It will not physically.它不会在身体上。
284Okay, that's not yet.好吧,这还没有。
285So remind him and then he has to stop it himself所以提醒他,然后他必须自己阻止它
286So you see how AI and human beings can work hand-in-hand together.所以,你可以看到AI和人类如何携手合作.
287It's not one or the other不是这样的
288So thank you very much for this now所以现在非常感谢你
289Probably as we look at how AI continues to evolve right you talked about we being really just at the也许当我们看AI如何 继续进化,正确的 你谈到我们真的只是
290At the at the beginning of this of this evolution and and we we very much agree with that我们非常同意
291So going for we say it can actually have a lot more potential now if you were to look forward所以,我们说,它实际上可以 更多的潜力现在,如果你是向前看
292What are the possibilities that you see being able to be unlocked by AI that really really excites you?你看到AI能解锁 真正让你兴奋的可能性是什么?
293I think我觉得
294In terms of the tick-tock product itself.在滴答产品本身方面.
295It's AI is really touching every part of our app experience我们的应用经验中 都有人工智能
296from creation the creation of videos从创建视频开始
297Right now, you know it because if you weren't a brand company现在,你知道,因为如果你不是品牌公司
298So you're selling a particular product and you want to post a video ad所以,你正在销售一个特定的产品 你想张贴一个视频广告
299It will probably cost you between a hundred to a thousand or even more可能要花你一百到一千块
300Dollars to make one video because you need to hire a team.做一个视频需要美元 因为你需要雇用一个团队。
301You need to hire a production house你需要雇个生产厂
302You even if you do it in house, you need to pay a salary即使你在家里做,你也要付薪水
303It will take you some cost to do that这样做要花点钱
304You can produce on our seed dance model你可以制作我们的种子舞蹈模型
305You can just put a your product picture in you can put in a video that you want your ad to look to look like你可以把一个你的产品图片 你可以把一个视频 你希望你的广告看起来像
306And we can produce a video for you, which is about one to two orders of magnitude orders of magnitude cheaper我们可以为你制作一个视频, 大约一到两个数量级 数量级更便宜
307Meaning, you know between one to ten dollars you can produce a video like this意思是,你知道一到十美元 你可以制作这样的视频
308So on the creation side you can see this所以在创作方面,你可以看到这个
309The short drama industry which短剧产业
310Really was an industry that we embraced first in China and then in the rest of the world真正意义上的产业,我们首先在中国,然后在世界其他地方
311Micro series, you know don't you if you haven't seen it that industry is now微型系列,你知道,如果你还没有看到 它,工业现在
312I think the major overwhelming majority of short dramas that being produced我认为绝大多数的短剧正在制作中
313AI produced using you know models like seed dance and other models.AI使用你了解的模型如种子舞和其他模型制作.
314So on the creation side you see this already happening所以在创作方面,你已经看到这种情况
315I I think is akin to opening a magazine today我想今天是开杂志的时候
316If you open a magazine a hundred percent of the photos are photoshopped, but you don't feel like they are如果你开个杂志 百分之百的照片都是被剪辑的 但你感觉不像
317CGI, you know, there's the human element of it and then there's the Photoshop to make it even betterCGI,你知道的,有它的人的因素 然后有照片店 使它更好
318We'll probably end you know in a state like this for video generation as well我们可能会结束 你知道在这样的状态 视频生成
319Where you have a lot of human creativity behind it and driving it but it's made a lot better with AI在那里你有很多人类的创造力 背后和驱动它,但它使很多 更好的AI
320So I think that part of creation is happening on the recommendation model itself所以我认为创作的一部分 发生在推荐模式本身
321We I know we our我知道我们自己
322Recommendation has always been based on machine learning but with the advent of these large language models建议总是基于机器学习,但随着这些大语言模型的出现
323I think in a lot of companies like us and others are realizing that from a fundamental technology point of view我认为,在很多公司中,像我们和其他人 正在认识到,从一个基本的技术观点
324You can make the models larger.你可以让模型变大。
325You can add in on a lot of new你可以加入许多新的
326Large language model capabilities into the traditional recommendation models and it's making it better and it's very obvious大型语言模型能力 进入传统的推荐模型 它正在使它更好 而且非常明显
327You know just by looking at the numbers that it really brings about a better user experience你知道,只是通过看数字 它真正带来 更好的用户体验
328there's a very large search a component to tick-tock as well and有一个非常大的搜索 组件的滴答以及
329We have building our own chatbot experience into that search because people don't just want to search and ask for a result我们在搜索中积累了我们自己的聊天人的经验 因为人们并不只是想要搜索和要求结果
330Sometimes they want to have a conversation about it有时他们想谈一下
331You know你知道吗
332What was that person wearing in the video and you have a whole conversation about it and that adoption is also in its double digits视频里那个穿什么衣服的人 你对此有一番对话 收养也是双数
333Within the app so it's and we talked about content moderation just now在应用程序里,所以它和我们谈论 内容节制刚才
334We talked about how our engineers are just using AI to code for them我们讨论过我们的工程师 如何利用人工智能为他们编码
335At least to help them code to a lot of the routine code today至少今天可以帮他们破解很多常规代码
336So you see this touching almost every single part of our business所以,你看到这个 触动几乎每一个部分 我们的业务
337But probably the most exciting if you ask me is what AI can do for science但是如果你问我 AI能为科学做什么 也许是最令人兴奋的
338Their bike dance there is a there's an effort to build他们的自行车舞是努力的
339the large language models as well and there's an effort to大型语言模型 也有一个努力
340See how we can collaborate with a lot of scientists to make some progress in science as well看看我们如何与许多科学家合作 也取得一些科学进展
341there's been some work that we've done in this area and我们在这方面做了一些工作,
342It is very exciting令人兴奋
343Yeah对
344I think there's a lot of a lot of people are seeing tremendous我觉得很多人都看到了
345Potential for AI to really push the boundaries of science for itAI真正推开科学界限的潜力
346And I think that's going to be very very meaningful for all of us.我认为这对我们所有人都非常有意义。
347I一、导 言
348agree约定
349I mean again from a lot of interactions we have of course as a bank with many of our customers and industry players我的意思是,从很多互动 我们当然是一个银行 我们的许多客户和行业参与者
350We do see that actually being deployed not just in some of things you talked about but in manufacturing in我们确实看到,实际上部署 不仅在一些事情 你谈论, 和制造
351Redesigning or reimagining manufacturing in redesigning and reimagining even pharmaceutical, right?重新设计或再设想制造 重新设计和再设想甚至制药,对不对?
352So in various various industries所以在不同行业
353We're actually very excited ourselves to see how the potential can can be can be unleashed even in our own industry in banking我们非常兴奋地看到 即使在我们自己的银行业 如何释放潜力
354we are completely transforming where you have something we call the我们完全在改变你有什么 我们称之为
355Operating model transformation.运营模式转型.
356We are effectively adopting AI in every step of the way how we are onboarding clients我们实际上在用人工智能 逐步接纳客户
357How怎么样
358We are putting in tools to help better with portfolio construction advisory services我们正在提供各种工具,帮助更好地提供一揽子建设咨询服务。
359Will allow you to execute your own transactions and at the end of the day also how to service how we try to improve both将允许您执行自己的交易 并在一天结束时 如何服务 如何我们尝试改善
360The customer journey as well as the journeys for all employees and it sounds like it's the same in the organization that you run now客户之旅 以及所有员工之旅 听起来你现在经营的组织也一样
361Building on that building on that given that there's so much potential there's already an adoption that we can see starting in various industries as well在此基础上,考虑到有如此多的潜力 已经有一个收养,我们可以看到 从各个行业开始
362What do you think it's still not receiving enough attention?你认为它还没有得到足够的关注?
363That we should really call out so that we can collectively pay more attention to I think you know the我们真的应该叫出来 这样我们就可以集体地 更多关注,我想你知道
364There actually is already a lot of attention to it.实际上,已经有很多人关注这个问题。
365But you know, obviously I think with anything that's so但你知道,很明显,我想 用任何这样的
366Exciting and creates so much upside.令人兴奋和创造出如此多的颠倒。
367I think Sushant described as dystopian utopian or as a paradox我认为苏尚特形容为 dystopian乌托邦 或者一个悖论
368I do think that you know, we do need to better understand, you know, how we manage the the risk that this could bring我认为你知道,我们确实需要 更好地理解,你知道,我们如何管理 风险,这会带来
369I think this the cyber security risks are quite real, you know我觉得网络安全风险是真实的
370I think a lot of us a lot of users may be giving我觉得我们很多人 可能有很多用户在给
371Permissions to the agents to do a lot of things for them without really understanding授权特工们帮他们做很多事情 却又不理解
372Without really understanding what kind of permissions you've actually given if you give your agent, you know a credit card没有真正理解,如果你给了你的经纪人什么类型的许可, 你知道信用卡
373Permission and you ask him to buy some stuff, you know, it could it could be overly eager to deliver允许你让他买些东西, 你知道,它可能太急于送货了
374You know what you want to buy and I'm buying much more than you need你知道你想买什么 我买的比你需要的多
375I think you know, there's a lot of我觉得你知道,有很多
376There's a lot of education that needs to happen还有很多教育需要进行
377But you know when the digital world first came out I grew up in an era where the internet was not a thing但你知道当数码世界诞生的时候 我成长于一个 互联网不是东西的时代
378you know when I was when I was你知道当我是 当我是
379In primary school it really came about when I was a teenager在小学的时候,我十几岁的时候就真的有了
380And I remember, you know, my my parents were terrified of buying anything online terrified, you know我记得,你知道,我的父母 害怕购买任何 网上惊恐,你知道
381Over time, I mean e-commerce is clearly a thing now.随着时间的推移,我的意思是 电子商务现在显然是一个事情。
382You have a robust payment infrastructure that the DBS is also part of您拥有强大的支付基础设施,DBS也是其中的一部分
383part of building globally全球建设的一部分
384so I think that these these things will come in time, but for the moment in time, you know, there's still a lot of所以,我认为这些东西会及时到来, 但是现在,你知道,还有很多
385Education that needs to happen on that sort of risk now on the corporate side现在公司方面需要面对这种风险的教育
386I think the risk of you know, cyber security is definitely real, you know, one of them one of the网络安全绝对是真实的 其中之一
387One of the sort of I guess unspoken, you know那种我猜没有说,你知道
388Knowledge within within the anti-spam industry is that really you cannot reduce it to zero反垃圾邮件行业内部的知识是,你实际上不能把它降到零
389What you're trying to do is make it more too expensive for people to do, you know too much spam on your platform你想做的是让人们更昂贵, 你知道太多垃圾邮件在你的平台上
390But you know with a lot of these但你知道很多这些
391agentic models the cost of just creating a lot of代理模型的成本 仅仅创建很多
392that kind of那种东西
393Behavior has fallen dramatically行为急剧下降
394So you need the right tools to now, you know defend against a fast-falling cost of所以,你需要正确的工具 现在,你知道防御 快速下降的成本
395Creating problems for platforms like us security problems for us.为我们这样的平台制造问题 为我们制造安全问题
396So having to build those I think is going to take a little bit of time所以,必须建造那些 我认为需要一点时间
397Thank you.谢谢
398So it sounds like again education so that we know how to use听起来又像是教育 这样我们才知道如何使用
399ai(单位:千美元)
400What to do what not to do?怎么办?
401That's actually reinforcing what show shared with us much earlier on at the same time from corporate angle pay special attention to cyber security as well这其实是强化了 早些时候与我们分享的节目 同时从公司的角度 特别关注网络安全
402I mean, it's a lot.我的意思是,这是很多。
403We have to train our employees not to uh, you know write我们要训练员工,不要写
404for example举例来说,
405Trade secrets onto a public chat bot, you know, you have to just tell them商业秘密在公共聊天机器人, 你知道,你只需要告诉他们
406This is not this is public and you have to tell them, you know这不是公开的 你得告诉他们
407There are certain privileged legal information that just you cannot this is not your lawyer, you know有些保密的法律信息... ...你不可能... ...这不是你的律师
408A chat bot is not really your lawyer.聊天机器人不是你的律师
409It doesn't give you legal, you know protection它不给你合法,你知道保护
410So there's a lot of that sort of, you know understanding所以有很多这样的,你知道的
411Uh because it's such a useful tool to have so people, you know, sometimes are over因为它是一个有用的工具 所以人,你知道,有时已经结束
412Enthusiastic about how they use it.对如何使用它充满了怀疑。
413So I think understanding, you know, where the boundaries of what is所以我认为理解,你知道, 什么是边界在哪里
414What is reasonable and acceptable use is important.合理和可接受的用途很重要。
415So maybe just所以也许只是
416As a wrap up as a final thing作为最后的结局
417There is a lot of talk obviously in more recent times about ai近来显然有很多关于ai的话题
418Being a threat to human beings because we are in a danger of it威胁人类,因为我们有危险
419Replacing or displacing human beings.置或迁人.
420What are your thoughts about this?你对此有什么想法?
421I mean we will see I think it may be a little bit too early to be uh to be too conclusive about this我的意思是,我们会看到,我觉得它可能有点太早 做呃太确定这一点
422You know on the one hand, I think we are the science the technology is not there yet你知道,一方面,我认为我们是科学 科技还没有到来
423The example that I hear a lot is our brain is 20 watts我听到很多的例子是我们的大脑是20瓦
424Our data center is in gigawatts我们的数据中心是千兆瓦
425Our brain can do things that you know, the model still cannot do my long-term memory retrieval for my brain is good我们的大脑可以做一些你知道的事情,模型仍然无法做我的长期记忆检索,因为我的大脑是好的
426I can go back to when I was a child我可以回到小时候
427I don't think you know the the large language models can do this easily at this moment in time我不认为你知道 大语言模型可以轻易做到这一点 在这个时候
428I can learn a lot visually我看得见很多东西
429I can look at something and learn a lot visually as a human being我可以看一些东西,学习很多 视觉作为一个人
430the models are still模型还在
431Well, I guess they're catching out, but they're still unable to learn certain things at my speed as a human being嗯,我猜他们正在赶上,但他们仍然不能 学习某些东西,以我的速度作为一个人
432so the the technology is evolving, but you you know技术在发展,但你知道
433You can see a world where?你可以看到一个世界在哪里?
434Silicon is just superior to carbon硅比碳强
435You can't see a world, you know, it will it will last longer你看不到一个世界,你知道, 它会持续更长的时间
436You know, it's more scalable.你知道,它更可伸缩。
437It's easier to connect, you know information together than us human beings connecting information together连接起来比较容易 你比人类更了解信息
438so it is a non-zero chance that uh, you know, it is something that所以,这是一个非零的机会,呃,你知道,这是一件事,
439could become可以变成
440Superior in certain aspects.在某些方面居上。
441Yeah对
442Thank you.谢谢
443I I do agree that that obviously as with anything we have seen technology evolve over many generations我同意,这显然就像我们所看到的 科技在几代人中的发展
444Technology is not new技术不是新的
445And even with the evolution of technology, we always still find human beings即使科技的发展,我们仍会发现人类
446Right at center being augmented being able to do bigger things more things faster things中心正在增强 能够做更多事情 更快的事情
447So I guess it is exactly why in dbs we say we want to be an ai-enabled bank所以,我想这正是为什么 在Dbs我们说 我们希望成为一个AI的银行
448But with a heart the approach we've always taken但用心,我们一直采取的方法
449Is that we believe it is not about what ai will do to human beings我们是否相信这不是关于 AI会如何对待人类
450But rather what ai will do with human beings and what ai will do for human beings但反之Ai会如何对待人 以及Ai会如何对待人
451Just like the technology advancements we have seen since the industrial revolution and beyond就像工业革命以来的技术进步一样
452On that note show.在那张纸条上
453Thank you very much.谢谢
454And thank you for your attention.谢谢你的注意
455Thank you.谢谢
456Thank you谢谢
该视频共有字幕 458 条。解锁更多字幕为会员功能,请移动到 价格