Name: Adobe 加入 AI 戰場，推出全新的 AI 工具「Firefly」 (Adobe unveils new AI tools)
Uploaded: 2024-04-02T21:01:10.000Z
Duration: 8 min

程度副詞

This week at the company's annual conference in Las Vegas, Adobe emphasized that AI is very much part of its future and that it plans to capitalize on this big technology shift.

本週在拉斯維加斯舉行的公司年會上，Adobe 強調，人工智能是其未來的重要組成部分，並計劃充分利用這一重大技術轉變。

基礎被動語態

Here with more, we are now joined by Adobe's David Wadwani.

Adobe 公司的大衛-瓦德瓦尼（David Wadwani）將為我們帶來更多精彩內容。

I wanna get right to a data point that stood out here.

我想直接問大衛，這裡有一個數據很突出。

Adobe now sees its total addressable market reaching 293 billion in 2027.

Adobe 現在認為，到 2027 年，其可尋址市場總額將達到 2930 億美元。

That is up from a projected $205 billion in 2024.

這比 2024 年預計的 2,050 億有所增長。

So, you know, David, that's a nice size jump.

所以，你知道，大衛，這是一個很好的跳躍。

What explains that jump and how much of that market do you think Adobe can ultimately capture?

你認為 Adobe 最終能佔領多少市場？

First of all, it is a significant growth in what we see in the market.

首先，我們在市場上看到顯著增長。

And a lot of it is driven by the fact that generative AI is helping make creativity and communication.

其中很大一部分原因是生成式人工智慧正在幫助創造和溝通。

And also marketing and digital experience is much more approachable for everyone from communicators, all the way up to the largest enterprises in the world.

而且行銷和數位體驗對於從傳播者到世界上最大的企業的每個人來說都更加平易近人。

And given that we already do business across that spectrum, it's giving us an opportunity to reach those customers in a much more significant way than we've been able to reach in the past.

考慮到我們已經在這一領域開展業務，這給了我們一個機會，以比過去更重要的方式接觸這些客戶。

So we feel that we are going to be one of the primary winners of that market.

我們認為我們將成為該市場的主要贏家之一。

And we're very excited about the announcements we had at Summit yesterday.

我們對昨天在峰會上宣佈的消息感到非常興奮。

Yeah, and so David, let's dig right into some of those announcements.

是的，那麼大衛，讓我們來深入瞭解其中的一些公告。

第一，你揭開了 Firefly 的服務。

Walk us through that product, David, and just what it's gonna mean for Adobe customers.

請向我們介紹一下該產品，以及它對 Adobe 客戶意味著什麼。

Great, yeah, for those viewers that don't know what Firefly is, Firefly is Adobe-owned generative models,

很好，是的，對於那些不知道 Firefly 是什麼的觀眾來說，Firefly 是 Adobe 旗下的生成模型，

so foundation models that let you create the widest set of AI creative content in the world, everything from imaging to design to text to vectors and other capabilities coming like video and audio and 3D later this year.

也就是基礎模型，可以讓你創建世界上最廣泛的人工智慧創意內容，從影像、設計、文本到矢量，應有盡有，今年晚些時候還將推出視頻、音頻和 3D 等其他功能。

What we did was we enabled all of the ability to reach all of that through API.

我們所做的就是通過應用程序接口（API）來實現所有這些功能。

So you can now, through an API, an application programming interface, enable automation in enterprise workflows so that you can generate content, you can assemble content,

你現在可以通過應用程序接口（API）實現企業工作流程的自動化，這樣你就可以生成內容、組合內容，

and then you can distribute that content through an automation layer that helps enterprises move much more quickly and produce more content that they can use for personalizing their experiences with their customers.

然後通過自動化層分發內容，從而幫助企業更快地移動和生成更多內容，用於為客戶提供個性化體驗。

And another product I wanna touch on with you, David, because it made some headlines, was GenStudio.

大衛，我還想跟你談談另一個產品，因為它上了一些頭條新聞，那就是 GenStudio。

So this kinda sounded like a sort of end-to-end platform for delivering campaign content, David.

大衛，這聽起來像是一個端到端平臺，用於提供活動內容。

Every enterprise is struggling with content velocity, so they need to think about their content as a content supply chain.

每家企業都在努力提高內容速度，他們需要將內容視為內容供應鏈。

How do you start by planning what you want to build?

如何開始規劃你想建造的東西？

How do you actually go through the process of creating that content?

如何實際完成創建內容的過程的？

How do you decide how to activate it and distribute it and what channels you want it to go out on?

如何決定如何刺激它、傳播它，以及希望通過哪些管道傳播它？

And all of it sort of sitting on a platform of how do you manage, just the proliferation of content you're creating.

所有這一切都建立在一個平臺上，你如何管理你正在創造的大量內容？

Digital content is really the way that enterprises need to engage their customers and drive and grow their business.

數位內容確實是企業吸引客戶、推動業務增長所需的方式。

And this is a single solution and application that they can do all of those five stages on.

這是一個單一的解決方案和應用程序，他們可以在上面完成所有這五個階段的工作。

You know, David, when we talk about AI and these sort of AI innovations you are announcing here, one question, certainly for investors, when are they gonna see meaningful AI monetization, David?

大衛，當我們談論人工智慧和你在這裡宣佈的這些人工智慧創新時，有一個問題，當然是針對投資者的，他們什麼時候才能看到有意義的人工智能貨幣化，大衛？

Yeah, we've been very clear with investors that we are hyper-focused on really the proliferation of the use of this technology, starting in October of last year when we took a lot of this technology generally available.

是的，我們一直向投資者明確表示，我們非常關注這項技術的普及應用，從去年 10 月開始，我們將這項技術的大量應用推廣開來。

And since then, we have seen an incredible adoption.

從那時起，我們看到了令人難以置信的採用率。

First of all, one of the key things about the way we do AI is we build these products, these models responsibly, so everyone knows what we train on.

首先，我們做人工智慧的關鍵之一是，我們以負責任的態度打造這些產品和模型，讓每個人都知道我們在訓練什麼。

And then secondly, we integrate it into our tools, everything from Photoshop to Adobe Express, all the way into our enterprise products that we just talked about.

其次，我們將其集成到我們的工具中，從 Photoshop 到 Adobe Express，一直到我們剛才談到的企業產品。

We've now generated over 6.5 billion assets with all of this, so we feel really good about that proliferation.

目前，我們已經創造了超過 65 億美元的資產，我們對這種擴散感到非常滿意。

And we are now starting to look at different ways to monetize it more actively.

現在，我們正開始積極探索不同的貨幣化途徑。

We've said we should expect to see some of the benefit of that in the back half of this year, and certainly ramping going forward.

我們已經說過，預計今年下半年就能看到其中的一些效益，而且肯定會不斷增加。

But we're thrilled with the adoption, and that gives us all the levers we need to monetize this over time.

但我們對採用率感到非常高興，這也為我們提供了長期貨幣化所需的所有槓桿。

David, when you talk about AI, one issue that comes up, and for good reason, it's copyright issues.

大衛，當你談到人工智慧時，有一個問題就會出現，這是有道理的，那就是版權問題。

What's interesting to note is Adobe actually provides legal protection for users, a legal shield.

值得注意的是，Adobe 實際上為用戶提供了法律保護，即法律盾牌。

Yeah, this is very important, and thank you for asking the question.

是的，這非常重要，謝謝你的提問。

我覺得這件事的相關報導還不夠。

The problem with AI is also a question and a discussion we need to have around: how is the AI trained?

人工智慧的問題也是我們需要討論的問題和討論：人工智慧是如何訓練的？

Where are we getting the content, and how are we enabling that content to be trained?

我們從哪裡獲取內容，又如何使這些內容得到訓練？

Adobe is one of the only companies that trains our models on content that we have 100% access to and 100% license to.

Adobe 是僅有的幾家對我們的模特進行內容訓練的公司之一，我們對這些內容擁有 100% 的訪問權和 100% 的許可權。

We're also very transparent about how we train that license.

我們對於如何尋練該許可證也非常透明。

So therefore, everything that people use for generative AI, if you're an organization, really we're one of the only choices in market.

如果你是一家企業，我們是市場上唯一的選擇之一。

The second thing we do is that everything that gets created with our artificial intelligence, or any of our tools for that matter, is inserted with content authenticity, meaning it's like a digital label, a nutrition label for how this content was created,

第二件事是，用我們的人工智慧或我們的任何工具創建的所有內容，都會插入內容真實性，這意味著它就像一個數字標籤，一個營養標籤，標明這些內容是如何創建的，

so that not only do enterprises know where they got the content, but when the content does go out, consumers will know how the content was produced.

這樣，不僅企業知道他們從哪裡獲得這些內容，而且當這些內容發佈出去時，消費者也會知道這些內容是如何產生的。

And we think this is a really important thing, especially in an election year when over 50% of the world is going out to vote.

我們認為這是一件非常重要的事情，尤其是在大選年，全世界超過 50%的人都會去投票。

David, I wanna bring this back to investors.

大衛，我想把話題拉回到投資者身上。

You know, when you talk about your TAM, David, that total addressable market, and you see it kind of growing mid-teens, but you also are just projecting about 10% growth.

你知道，當你談到你的 TAM 時，大衛，即總的可尋址市場，你看到它在中位數增長，但你也只是預測 10%左右的增長。

There's some investors, David, who might hear that and think, you know, Adobe is a real leader here. It should be growing faster than the overall market.

有些投資者聽到這些可能會想，Adobe 在這裡是真正的領導者，它應該比整個市場增長得更快。

I come back to, you know, Adobe has always taken a long view on the market, and that's really what made us the company that we are today.

Adobe 總是從長遠的角度看待市場，這也是我們成為今天的公司的真正原因。

We could, of course, take actions that are gonna drive more revenue in the short term, but we feel the right game to do right now is to drive proliferation, really get everyone adopting our technology, and then ramp the value and the monetization over time.

當然，我們也可以採取一些措施，在短期內增加收入，但我們認為現在最應該做的是推動技術的普及和使用，真正讓每個人都採用我們的技術，然後隨著時間的推移提升價值和貨幣化。

For us, it is about making sure that Adobe and Firefly and GenStudio are the places that companies and individuals come to create content more effectively, and then monetization will follow.

對我們來說，要確保 Adobe、Firefly 和 GenStudio 成為公司和個人更有效地創建內容的地方，然後再實現貨幣化。

And David, I wanna get you out of here on a question about competition.

大衛，我想讓你回答一個關於競爭的問題。

It interests me when folks saw OpenAI's Sora, David.

當人們看到 OpenAI 的 Sora時，這讓我很感興趣。

I'm sure some Adobe investors, they might've gotten nervous, right?

我相信一些 Adobe 的投資者，他們可能會感到緊張，對吧？

I'm more interested in how you at Adobe saw that, and whether actually you saw that as a kind of opportunity, meaning it's one thing to generate an image, but, you know, if you wanna cut, you wanna edit, et cetera, you still have to look for those Adobe tools.

我更感興趣的是，你們 Adobe 公司是如何看待這一點的，以及你們是否真的將其視為一種機遇，也就是說，生成影像是一回事，但如果你想剪切、編輯等，你仍然必須尋找那些 Adobe 工具。

Content production has always had multiple steps.

It starts with capturing content, then it goes into producing that content, and it goes into distributing that content.

首先是捕捉內容，然後是製作內容，最後是分發內容。

Adobe has always played in the producing phase of this, and we think that's a massive market that we're gonna continue to play in.

Adobe 一直在生產階段發揮著作用，我們認為這是一個巨大的市場，我們將繼續發揮它的作用。

What Sora does is it really addresses the capture part.

Sora 所做的就是真正解決捕捉的問題。

So today, most of the content that's created that gets edited in our tools is through cameras and video machines.

今天，我們工具中編輯的大部分內容都是通過攝影機和影片機器製作的。

Today with Sora, we're gonna expect to see more video created that needs to flow into our tools, and we're gonna integrate Sora and other video models directly into our tools.

如今，有了 Sora，我們有望看到更多的影片被創造出來，需要流入我們的工具，我們將把 Sora 和其他視頻模型直接整合到我們的工具中。

I do wanna also add that we have our own technology.

我還想補充一點，我們有自己的技術。

The research is very similar to what Sora is doing, and we expect to have that out later this year.

這項研究與 Sora 正在進行的研究非常相似，我們希望在今年晚些時候完成這項研究。

So we feel very pleased with what Sora is doing.

我們對 Sora 所做的一切感到非常滿意。

We'll have our own with hopefully increased levels of controllability, but it all flows into our tools and helps our business.

我們將擁有自己的工具，並希望能提高可控性，但這一切都會流入我們的工具，並有助於我們的業務。

David, thank you so much for joining the show today.

大衛，非常感謝你參加今天的節目。