Future & Remix 论坛回放 1 | The Evolution of AI in Creative Work
39:53 Share

用 声音 碰撞 世界 生动活泼。 Hello, 大家好， 欢迎来到 我们 今天 的 what next 科技 早知道， 我是 丁 降。 大家 可能 也 知道 在 过去 的 几个 月 时间 内， 我们 生动活泼 的 小伙伴 们 一直 在 忙 一个 活动， 叫做 future and remix。 这一次 的 活动 我们 把 它的 名字 叫做 A I 如何 重塑 我们的 内容 创作。 这样 一个 其实 汇集 了 很多 很 前沿 的 A I 创作 内容 领域 的 创作者， 以及 很多 的 科技 行业 领域 的 从业者。 这个 活动 我们 其实 是 跟 ultra 香港 音乐节 一起 合作 联合 举办 的。 终于 在 上周五 在 香港 完成 顺利 的 结束了。 结束 以后 我们 也 收到 了 很多 播 的 现场 的， 还有 参会者 的 给 我们的 正向 反馈。 所以 在 今天 的这 几期 节目 当中， 我们 也 希望能够 把 会议 上面 的 这些 精华 跟 大家一起来 分享 一下。

像 上次 和 俊 宇 在 节目 中 聊到 的 一样， A I 它 其实 就是 一种 utility， 它 就是 一种 基础设施。 所以 我 发现 其实 最近 在 我 跟 很多 科技 圈 朋友们 聊天 的 过程 当中， 大家 其实 对 A I 如何 落地 应用， 其实 还是 有 一点点 灰心 和 丧气 的那 我 希望能够 通过 这 样子 的 一次 跨行业， 甚至 是 跨地区 这样的 一种 交流， 能够 产生 出 一些 非 共识 的 想法， 产生 出 一些 灵感 和 火花。 这一次 虽然 是一个 闭门 的 会议， 然后 大家 其实 也 不用担心， 我们 把 所有的 节目 完整 的 编辑 成了 三期 节目。 然后 今天 大家 听到 是 我们的 这个 音频 的 节目， 下周 我们会 在 youtube spotify 以及 我自己 的 小红 书 上面 看到 我们 制作 的 视频。 因为 这次 其实 是一个 全英文 的 节目， 我们 当然 也会 配 上 我们的 中文 的 翻译。 然后 包括 台上 的 一些 嘉宾 也 分享 了 他们 用 A I 制作 的 各种 视音频 等等 的 这样的 一些 素材。 我们 也会 把 这些 内容 完整 的 放到 我们的 视频 当中。 所以 期待 大家 下周 在 不同 的 平台 上面 关注 我们的 视频 内容。

So thank you.

第一个 panel 也是 我自己 主持 的 一个 panel， 主题 就是 A I 对 内容 创作者 的 影响。 我们 也 分别 邀请 到了 几位 其实 前年 已经 在 用 A I 创作 的 一些 创作者， 然后 很多人 他们 已经 获得 了 国际 上面 的 大奖。 比如说 是 与 非常 多 知名 艺人 合作 过 的 A I 导演 朱 尼， 以及 受邀 参加 这次 ultra 音乐节 的 创作人 lona， 还有 V R 演唱会 的 制作 公司 amaze， V R 的 C E O Steve ly。 他们 都是 大量 使用 A I 工具， 并且 将 A I 整合 到 日常 的 工作流 里面 的 创作者。 另外 我们 也 邀请 到了 香港浸会大学 艺术学院 的 教授 john y pun， 来 分享 他在 艺术教育 方面 的 一些 前沿 的 探索 和 他的 研究。

令 我 印象 深刻 的 是啊， 这 几位 嘉宾 其实 都是 来自 不同 的 创作 领域。 但 他们 基本 都 提到 了 他们 利用 不同 的 A I 工具， 大大 提升 了 他们 在 这个 工作 上面 的 效率， 包括 结 省 了 很多 他们的 时间 和 人力 的 成本。 比如 朱 你 就说 到 在 传统 的 电影 制作 里面， 由于 时间 和 经费 的 限制， 很多 的 特效， 比如 像是 这种 燃烧、 爆炸 等等 一些 场面， 一两次 就 必须 完成。 但是 A I 的 使用 能 让 这些 原本 复杂 的 制作 摆脱 了 时间 和 空间 的 限制， 所以 他 也就 释放 了 很多 的 创作 的 灵感， 创作 的 空间。 嘉宾 Steve 他 也 提到 了 他们 公司 利用 ai 将 整个 的 V R 演唱会 的 制作， 大大 提升 了 他们的 效率。 比如 在 过去， 他们 基本上 会 用 40到50个 人的 工作量， 可能 是 需要 几个 月 的 时间 才能 做 完 一场 演唱会。 但 现在 只需要 用 两三个 人， 基本上 几天 就能 完成 了。

最后 我们 其实 还 探讨 了 关于 A I 生成 内容 的 一些 争议。 比如说 A I 会不会 取代 创作者， 会不会 让 作品 失去 了 它 本身 的 一些 灵魂 或者 是 意义。 Luna 其实 就说 她 并没有 这样的 一些 担忧。 因为 现在 的 A I 的 普及， 其实 是 让 更多 的 人 有了 实现 自己的 想法 的 工具， 让 这个 实现 的 过程 更加 的 容易 和 更加 的 高效。 以前 比如说 是 需要 一整 支乐队 才能 完成 的 一首 歌曲， 现在 一个人 就可以 完成 了。 其实 这也是 让 更多 的 人 有 机会 成为 创作者。 另外 露娜 也会 说， 她 其实 是 让 A I 为 她 自己 提供 一些 demo 的 不同 的 一些 灵感 和 想法。 但 最后 怎么样 去 使用 这些 不同 的 片段， 然后 怎么样 整合 起来， 其实 是 由 他 来决定 的那 这首歌 依然 是 注入 了 他 自己的 品味 和 自己的 灵魂 的那 在 节目 正式 开始 之前， 我 还是 要 感谢 一下 我们的 合作伙伴 伙伴。 这一次 的 联合 发起方 z bra lab ultra hong kong， 然后 还有 我们的 支持 单位 香港 投资 推广 署， 包括 我们的 赞助 方 图拉斯 以及 鹰牌 花旗参。 好了， 下面 就是 我们 今天 的 节目。

So this panel, we're gonna discuss pretty much all about how creators using A I, and then they are all the four runner for using A I. And there are creative process. And then a lot of them already have feedback, what clients paying for their AI embedded work.

Yeah so i'm really looking forward to this panel. Uh so maybe um first of all, let's start with um maybe june. When did you start using a in your creative work and then .

how do you use them? Hello everyone, i'm Jenny and I am a AI director and production designer are using cutting edge technology to immediate production to uh gold brands and top celebrity which including fashion field m music video and advertising production and I also uh do some research in virtual identity and engage in critical thinking about digital humanity.

Hi everyone is lina. Uh, I am a music CER, also A D J. Am a song writer.

I studied classical music. I measured in conversation and sound design while I was in the U. S. And after I came back, I signed with universal music group work as a producer.

So my job is a right songs for the pop artist rappers and I am also um a singer myself. So I um yeah use a lot of A I A in my work as well. That means stream stuff like snow video sound design tools like ten diffusion. I'm not if guys .

ah hello my name is is my Steve. I'm a co founder and see about a mage V R A major V R is a fear concert. Platon company if you found IT in two thousand fifteen.

So it's been ten years and yeah we found the in the company. Now we are moving to the L A. And then we also have the subsidy in south korea.

I would just start my career as soft engineer twenty years ago. So i'll say i'm the quite engineer engineer and I start to work at the spinning company and like banning company, the major concert. And I work at the which is kind of we chat like service in south korea.

And like after that, I believe in A I V R, I believe vrar has said would replace the T, V and monitors as a process display. I thought apple really list vir has said two thousand seventeen for, unfortunately this this year though so but now is getting Better. And then we work, as I mention, we are a concert, the company, which means we are working with music.

So we started work with the U. S arts like Megane tipping, and then we expand to the kip hop ah so we started to work with the aspa and X O ky, and the recently we work with the tomorrow together, which is the big to music artist. Now we are actually launched in the movie theater.

Is not many people, half of your hazard yet. So we set a ticket to the to the fence. And then when fan comes to the movie theater, fear has said would be retold, and they can his peers of via concert with the movie theory system.

So we now lunch in the south korea and the U. S. And japan and that we are discussing other countries.

I don't know you guys know the V R production, but the V R, we need to fill out all the space is which is there's a lots of. So if we just followed like trade film production of the game production, IT requires a lots of menu work. So we use A I for automate menu works. So that's what .

we are going. Professor pn.

um i'm john y and I am a educated or nine adec as so now I believe little different from the other panellist uh i'm in education. I'm also um i'm the founding thing of the school with film program, with music program, with visual program and also an art technology program plus a global a business degree in global entertainment.

So ah we have about eighteen hundred students, uh we employed about in the two hundred the professional uh artists and uh uh filmmakers and looking for us. So uh we basically the only university in this part of the world with these are elements together. My other role is uh as a uh association p of interdependent enable research. I am by heart at a scholar. H I know love doing research, and I also love bringing people together from different field.

So I look after six research al apps, all related to A I of mental creativity, no a data economy, no computation, no medicine, health and well being and also the style and then also several centralized A I related facility one of those um which I am the director is a more cap, a research and visualization that wishes the only one in this part of the now we have a one of the most accurate of precise a lab that um we can capture multiple actors within the space that that will allow up to five people with the accuracy that is so precise that will allow us to use the data for machine learning. And later I will explain some of the project we have done last but not at least in the i'm also musician and so i'm much are trying as A A orchestral conductor by nature a conductor you know a organize so I feel like no, i'm a project manager in so many of us, including, uh, uh, advocating for new technology. How did I get into the A I? I actually have always been interested in crossing over now are using digital technology.

Uh, uh, no, that was before twenty, twenty, you know. So no, I have always been using all out of tools and insure to create amazing. But then during the pandemic now we was all stuck in hong kong. And so my research partner in the l who's been A A, A data scientist who has been a really big a love of music. And I also love A I, so we work together and we got the first ever art technology funding research funding from the hong gong government in to six uh sixty million dollars to develop new A I technology platform that will transform how people create, communicate and how people receive and participate in the arts.

Officer has a couple really cool video. So that's done in A I research. We can I take a look.

Let's start with the other one first. This was down two thousand and twenty two. That was in home court.

If you remember, that was during the height of pandemic. We cannot have any quiet on stage in their performance. So we actually devote about all algorithms. M, now you hearing is A A I quiet singing live within seven and the orchestra.

But the real advancement here is you hear what language they were singing IT 我 向 南流。 到 香江 天安 的 海。 So that was during the time when is not possible to put a quiet on stage when everybody was still wearing mars.

I want to have a concert, knew I can put all the musician violin is all in mass, but I can't have acquired simply in a small space like that. So we ask out a research team the question, can you create a quire for me, where they can perform with me to life, you know, in real time? No, this is not playing a recording.

This is actually A, I quiet responding to my movement and uh so that was actually done by extracting, know the male voices were a world class. Uh, a solo is like domain o, say kalas. And the female singer was maria colors.

Maria colors was long dead now, I know of thirty years ago we extract her voice quality and then taught her to sing in putin water. And so, you know, that was a lot of, uh, uh, uh, try and error initially is sound terrible IT sounds of like somebody just know told me out of control is like a something come out of a horror movie. And then, but then slowly, we improve our algorithms.

M, we keep changing. Until there was one day, I remember when we were listening to one of a an example. And then we all look, look at each other and say, did you teach you to think like that? Did you teach IT to phrase like that? IT almost have their human quality do IT.

So, you know, that was a part of the process. Know how we saw that we're using A I in life performances. And uh, we early, I talk about the motion capture lab we created with forty cameras and but at the same time, um yeah this is A A more recent work this year.

In fact, I took a team, researched team to gracy and we are want to a do a piece of music and there was written based on Melody from the alzo o but the thing is that this music was actually really a long time ago, in the fifties, uh, you know, it's very westernize, but we want to find that color and then we want to see what is, uh, h looks like. No, in in this, a original setting. So we went to the auto setting, and that we also bring our science team to do motion capture of dance data, and that we taught our message to do IT like because of time.

I only give a snapp shot. I want to assure you, this is not animation. This is no completely AI generated, a visual that only using word proms.

So what we were able to do is we call out teg motion 啊 GPT。 So no. Now, with languish model, you have all of the language, no data, but is no.

What's lacking is motion data. So we were trying to do is to create this, uh, uh, uh, algorithms. M, but we could just type and say, OK, no. Act, no.

Walk like chaplin? no. And they go from this of side of the room to another side and then do a flip. And so you know, we we we created using in the the local language model with the small uh model that we collected. And then we combined you and then create or alga with them to create this kind entirely text, prompt a visual.

And then of course, I know that's a lot of post production engineer because initially, though, those hand was just so horrible, they looks like gilla those skills. You know, they have pretty faces where their hand was like this. And so so you know that's a lot of what we continue to improve so that .

baseball what we do with this come back to the stage. What do you guys think it's pretty amazing? Um maybe the next question now was um thinking about so professor is said that was kind of a tool that he used for his research and uh creation of the arts. Maybe jy can tell us more about when you create your M B his with uh all different celebrities, do you have any preferences? You see their drawbacks kind of where are you are using?

I will mention like some of the amazing too I have used about, which including the gi, which could 希望 in then also another two is one way and ChatGPT rumor and many others。 So because every different tools have different advantage. And this advantage, so IT is for me like making an A I movie, just like mixing and transitioning different tools.

And also, I think that's a lot of possibility and potential in the A I R creation, which are waiting for the artist to be exported. Uh, so my past are experience with all the different AI, too. So I just realized that this kind of news technology are giving birth to an entirely new art form and vision exception. So ourself, as the humanity has transition from a time where creativity was slowly driven by human brand, to a new century where creativity shape by the interconnected calibration between the human brand and intelligent brand have been.

the tag has improved. You are in every .

one is a huge we have actually A.

A, A listener have a question, did they feel like is probably the scaling law is not scaling as much out nowadays? So you don't think so, sorry. So you think the tag is still progressive?

Yeah a lot, a lot OK. And I want to say the competition between all the A I tech company is so so instance yeah and wear what what do you .

mean at that? Um the competition between artists, a lot of .

arts yeah because artist is trying to use A I to make the latest was so every day you will see a fend up forms and also defend meaning behind their art so is truly wonderful for me. And also we can see there's a lots of limitation about A I Y now, but that many artists is using the limitation of A I to create the amazing and wonderful work. And this is shaping the a section of human to see the A I. So uh, working with A I is more interesting for me to working with him .

like traditional tools. Um so lesson, when we talk off flying, you mentioned there's a lot of celebrities want to work with you and they want there's so interesting eye they know all about GPT or all different kinds tools. Can you share more about .

those things? I would say the the communication between the celebrity is is different because they have their own position about A I film making. Also, their perception is come from their traditional film making.

So the communication sometime is more like a battle. What kind of battles you guys are, because in traditional film making, they they do the exposure, they fired different kind of fire thing. And IT is so intense because there is no stuff behind the this kind of they feel making element. But in the A I, then we do the all the fire and explosion with A I with generative. If you working with A I, you input the pump, then the model respond and you can try many time, but in traditional fuel making, you can only try a few time based on the budget because is so expensive.

So I don't see worth the disagreement come from isn't .

no because they they will use the their understanding of the the real time to ask us to do the the similar thing in A I. But this is not the A I work because A S advantage is in presenting a lots of possibilities and you can try a lots of possibility in a really short term. So you can't just lost A I to do that exact thing in your mind. I see existing how A I was so .

probably not fill into their old workflow, me to create a new one that kind of the disagreement come from, know what about you? Like do you use AI in your creative process? Or is a tool to not consistent? And for your work, what do you use IT for?

I, I want to say I really like professors first work. So I use a lot of a virtual singer as well.

When you call IT virtual singer has there are a lot of A I tools, things like A A C studio, a studio and like synthetic because uh if your song writer but you are best singer and you're trying to like produce a demo for your client, it's actually hard because um if you can sing, you need to pay for a professional singer, probably need a like a good studio or even the engineer professional engineer that means you need a lot of time to put things together and need to pay a lot of money. But now it's like a virtual singer. You just need to sing IT IT and you can replace IT with a gracing er just like professors for work.

So A I is actually um changed a lot of way. The people create music. And the other thing that I use is a woodie like I guess a lot of people already know what IT is something like me journey, but it's a take A I tool um you can just input like a emotion and jana, you are just generate music in five minutes.

So if you're not a professional producer or song either, you can still write a song um but for professional musicians or producers, I think it's still helpful because when you lack inspiration, you can update something we call the audio clip like five seconds to ten seconds uh music clip to and then you can continue writing for you as so you can generate like unlimited version of a song so you can give you possibilities fast and like inspirit of vest instead of just drinking beer, struggling and stuff. Oh the other thing that I use is cod dance diffusion because um most of the time notice I produce edm music, electronic music. So the most important things in electronic music is a sound design.

So traditionally you will have to learn how to use this synthesizer or max for life, things you that you need to learn for three, four years or even longer. But now with A I technology with dense division, all you need to do is collect the sound you like, like water sound, pass sound, or like Allan workers lead anything to a folder and an upload to google drive. You can generate new sounds that the A I think you are like, uh, like for you, so you can use IT in your production. No.

I have a question. Just um your fan or people who listen to notice anything that you created by, I tell .

them me about this. I yes yeah but i'm not against A I had all I think uh some sound right to stating uh, video and so no tools like that might take away some people's job but I don't think you will cause the sound quality uh, fortunately, it's not good enough for commercial releases yeah maybe in a clear .

I don't .

think this can like actually because the database is not like high quality, it's hard to analyze all the data as at the end of the day, you need mixing engineer, master. They can only like help the process of. Creating music, but you cannot totally replace IT. Also some people to think, oh, you just generate a lot of stuff and you pick the parts of a sooner in would you give you you're not really writing and creating music, but me personally, I think picking the parts you like and fixing the things you don't like is also a part of creating cause by dead you are kind of choosing the things that will fit into your taste so I think helpful for a lot of people. Um if it's good or is the art is the good music that depends but I will say it's just A I creating because you are generating, not A I generating for himself.

not I doing that. And just like forty, maybe a Steve can tell a little bit more about you said that you pretty much increase ed your efficiency when creating a uh, V R concert. So you originally probably use like big teen for like months of work. Now I cut down to a few hours. Maybe you can tell little bit more about that.

Uh, so we are production actually we capture at A K at sixty a frame per seconds. And Normally a films is four k twenty four frame per second. And if you compare four k for S A K there are four times of the work.

And if you compare two, five from frame prosecutions to the sixty is two point five times. So every two years there are ten times more work for the V, R. Consol productions. And then we produced minutes of the the cause, or which is is kind of four hundred minutes of the film productions, but actually want more. That's because in V, R, like we can control where people looking at, which means like every the frame should be perfect.

If you think about the like iron man, the films when we close up the Robert down genia, like the background is black, which missed, we do need to render that high quality image for the the background for V, R. We need to render every frame in the highest definition. So we really need millions million dollars for the production, for the fear cause productions.

But as if we are mess up is not there yet. It's not possible to put, if you are concert economically. So we need to use the A, I, so that we are using the A I, so we automate the king out process, which means the king out. The live action r footage from the Green sky background IT requires less of composition ors, but we only have two people and that they provide some data and our property A M module doing the rest of the work. So and then we are using the A I for the tracking the dancers.

So we, we, we used to use animator to make the dance movement, but but now we just need captured the dancers in the cameras and R A M modules, make that things of movement to buy the themselves, so that we automate the many process, so we allow us to push of your concert sky. So like I would say for the competitor, like I think with the similar production, we might need to have like forty, fifty people, but now we have only two people. So I think this yeah can show the how A I can help the like productions. Officially.

you were also mentioned to me that direct like the next generation of interactive prada, they were are thinking about, but maybe in the a few years, maybe you can tell a little bit more about that.

Like replace some interactions for you like to keep up the five members for the two by together. So we allow people to choose their favor members. And then when they choose IT, let me show the some footage of the death favor members.

Because in movie theaters, they used to their dedicate person display. So which means we don't need to show the same thing to the older people. But we have that comes small interactions, but not just like the images, but maybe the three models.

So if we like, render the three model the artist in this high definition with A I, which I think we will happen at some point. Then we get a lot of people to look on this vehicle or IT get closure. So I say at the moment, there's last of you that we apply so but you will require a lot of time though because but I think the a is getting so fast that I can they expect the development. So i'm not fully sure at the three to five years, but we can see.

So i'm not sure ever i've seen this japanese meta video yeah, a lot people think is pretty creepy and is kind of hit people's and can value a little bit. June told me the decorators, your friend, yeah, is that kind of a reaction from people expected? Or how does he react after those late? Some criticism, sure. And then but madona pay for IT. That's huge win already.

I am also co funding and A I production house, which launching soon in this september and in our team there are artists from OpenAI, sora and A X T Z. So because we all see this, this thing is coming, and IT is coming so soon, so we just want to put all our effort on IT. So decline.

They love this kind of AI created work, really like they are going to care you to traditional creation process. So what what's the advantage for A I creative advertisement verses old when they are just like buzz. What is the quicker or cheaper?

What do you think? Um I would say because in my past experience with many A I feel I haven't come accost many hate from their audience yeah who watched my AI art but I know that many people when they first think of A I art, the first thing that come to their mind is IT is not out IT IT doesn't have a soul and also that many people focus on its potential to produce force and trip content.

But for me, I have a totally the opposite will about that because most of the feedback come from my A I feel has been overwhelmed positive. So I think the reason why is that because audio recognize the unique and severe artic version that the day, and I also believe that this will shape the content of the next generation, because I would say because A I have enable artists from different background to express their idea more quickly. And as a so the the really the only thing that truly matter is how powerful this idea and whether they can move and inspire the audience.

So for me, every creative attend is part of a continue feedback loop. The faster the the artist can generate, the more idea that the artist can explore. So leading the world that are uh touch people.

So at the same time, I would say audience is also seeking for higher quality content about which is different from the traditional filmmaking because I would say traditional filmmaking have the limit. So this is IT is the thing that rising the standard for the creator. So I I also believe that anger for new experience, content and offer something they have never seen and her before and experience before.

Uh, maybe we don't have much time. Maybe we will want to like the future. So maybe the first question should be addressed to a professor pan, do you use like A I tool for you ah when you educating your students? And that, is that more easy? Or what do they, uh, do novaes like? Yes.

no, no. Quit question. In fact, no, we are. The first university did a job ChatGPT.

Other city university was the first one initial ended. You remember the time? Yeah, all the universe are. Now, we are not gna allowed student to cheat, you know, with chapt. But we were the first unity.

First is say, we need to show them how to use that tool and that responsibly and also how this can go create you know new opportunities for them. Um just a the other day um I got this description from um my one of my technician. I also teach my staff to take advantage of you that AI technology.

The other day I called them you a little was then I say, no, it's description OK for another show we get to do. And I say, oh, you've a very good writer. Did you write? I got a happy face. He did a very good job prompting, you know, giving no excEllent idea for the a the ChatGPT to produce a contact that exactly what we need.

And that in the past, you know, I swear L A lot of time editing somebody else writing, but now i'm very pleased, knew they have good work and they have the good wisdom to put the right prom to shape, and do IT reiterate again and again until they get something day, think is good enough. And then samit, to me, I think that was good. So I think, in a way, know that's gonna be more and more opportunity, divorce or go creates.

Do this do have to learn instrument? Maybe this also came the dress. Slone's learning instrument is super hard.

right? Yes, they still do. Or in fact, lona was everything I would say, uh, no, we get the idea from the AI, but you do execute.

And now there are still a lot of gaps. Now, for example, with human voices, you are very right. You know, those in the voices that we use for more cup, sometimes it's not really usable.

But now I recently have I saw the company that we are looking at doing walkup singest. yeah. But the pain point right now is now getting very good. Yeah, yes, yeah. But that is still know.

That's why that's the little area where looking at think is the completely change film for the national grows a real orchestra and that we build on top of that. But you will never replace human being, but I think I will replace human being who don't use the A I too。 Yes, yes.

very well. thanks.

yes. So the other day um you remember a dinner ask me, do you think um people were actually like A A music? And I think about IT because, uh, I have a small company that my partner, you can, we are the provider to doing that.

We create A I clipsed. It's like shock clipsed only like fifteen seconds. The reason why we do IT is to test out if it's gonna be a hit course or a hit hook that will be widely used by the users. So like that in the days the record companies, they spend a lot of time and money into like a full production.

But now what they do is to purchase a lot of short clipsed, which A I generated, and then test out in doing to sea if it's gonna be a hit because, uh, the music promotion way for us nowadays is not like long time ago, only through T, V. And radios anymore, but shot videos like growing tiktok and youtube shot videos. That's why I think the fifteen second clipsed of music is crucial because if it's good, people will, sir, to a music or spotify.

On like apple music or QQ nest. But if it's not good, your music hard to be heard because there are so many content these days. So if for doing now, it's like if the clip um are widely used by the users in doing and they will have the money invest into to h producing the full full length music and everything. So I think why people were like the the content that A I generate, the A I make, it's proven by the you know the the algorithm and in stuff but is a good music again, i'm not sure. But I think people will like kids because she's hard to tell that after if you make a woman year, you don't know if all the money put into the promotion is gonna a hit.

I think the audience only care about good content and new experience. They doesn't care much about whether this and with A I or traditional fum OK. But we were just .

mentioned earlier, some people would think this is great by A I, this is no soul. And then .

there's like 鄙视 链。

Contact change you are by are probably not good, but this will change, right? You guys pretty positive this yeah. All right.

Thank you. Thank you for your question. Thank you for our panel. That is awesome discussion. Yes, please keep them around a post.

节目 的 最后 也 和 大家 分享 一则 我们的 生动 胡同 会员 的 一个 消息。 九月份 正值 开学 季， 我们 希望能够 让 我们的 学生 朋友们 更加 便宜 的 成为 我们的 会员。 所以 从 去年 开始， 我们 就 增加 一个 看起来 不怎么 合理， 但是 大家 都 很爱 的 一个 教育 折扣。 正常 加入 我们 互通 会员 其实 是 365元1年。 如果 使用 我们的 教育 折扣， 其实 是 120元， 其实 是 将近 3折的 这样 一个 优惠 力度。 成为 我们的 会员 可以 解锁 我们 旗下 所有的 付费 节目， 包括 不值 金钱， 还有 跳进 兔子洞， 同时 还 可以 收到 我们 每 周一封 的 邮件 通讯， 就是 new sector 里面 记录 了 我们 制作 内容 的 幕后 故事， 然后 以及 我们 当下 的 所思 所 想。 点 开 我们的 说 no， 也就是 节目 详情 中的 文章 链接， 可以 看到 我们 为什么 设计 这样的 一个 特别 有 力度 的 一个 折扣 给 我们的 学生 小伙伴 们。 然后 也会 有 我们 生动活泼 的 小伙伴 们 对 自己 18岁说 的 一句话， 欢迎 各位 来 订阅， 也 欢迎 大家 把 这样的 一个 特别 有 力度 的 折扣 分享 给 身边 正在 上学 的 小伙伴 们。