Is AI The Adobe Killer?

Summary:

• AI, or artificial intelligence, is the buzzword in today’s financial markets.
• The disruption threat for many companies is enormous, software in particular.
• How does AI impact Adobe Inc.? Is it a threat to the business model?

Introduction

There’s currently a lot of discussion around artificial intelligence, or shorter, AI, and the potential implications it might have for almost any business. Generative AI models like ChatGPT, Dall-E and Midjourney look very impressive even though we are in the early innings.

The exponential rise of AI has even led some experts to sign a document that calls for a slowdown in the development of this technology. On the other hand, countries such as Italy have chosen to ban ChatGPT entirely, fearing its impact on the economy.

AI will undoubtedly have considerable implications for many companies, but we don’t think it’s a “do or die” moment for all. Investors have grown increasingly worried about AI’s impact on tech-oriented businesses such as software. Even though we understand why this is the case, we don’t think AI will put every software company out of business.

In this article, we discuss the Adobe Inc. (NASDAQ:ADBE) case. Can Adobe businesses survive AI or maybe even benefit from it?

AI’s impact on Adobe’s business

We are Adobe shareholders, but have grown increasingly skeptical of Adobe’s competitive position over the last few months.

This has been the case mainly for two reasons. Firstly, the company’s $20 billion acquisition of Figma made us question our thesis. Companies with strong moats don’t have to pay 20 times sales to acquire a competitor. We covered this topic in our article “Why Adobe Is Buying Figma for $20B,” so we won’t go into detail here.

Secondly, the rise of AI made us question the value added by Adobe’s tools. Part of our Adobe investment thesis relied on high switching costs. The rationale was that many professional designers had “grown” with Adobe’s tools, so these professionals would remain committed to them due to a steep learning curve. Adobe’s creative tools have been the standard in the industry for many years, and they are not easy to use. We believed both conditions made the appearance of a serious incumbent a low-probability event.

With the rise of AI, not only have Adobe’s design programs gotten simpler (thereby reducing the switching costs), but there has also been an influx of new entrants into the space. One such entrant is Canva, an Australian company founded in 2012. Canva has been very successful among non-professionals due to its ease of use and great user interface. The professional market has not been Canva’s bread and butter, but AI is helping the program improve and now allows users to get an almost professional-looking design in a matter of minutes. Adobe’s response has been Adobe Express, which is already seeing significant traction in the mass market.

After these AI-aided programs came generative AI. We imagine that many of you have already seen what generative AI programs such as Dall-E or Midjourney are capable of. Basically, a user can write a text prompt, and the AI will generate an image from scratch. For example, we typed “a cactus riding a wave” in Dall-E, and this is what we got:

The result is far from perfect, but it took us 10 seconds to get the image. If we had started by looking at a blank piece of paper, it would’ve probably taken a couple of hours. The productivity gains are evident. We think this will have negative implications for Adobe’s non-professional business, but after doing some expert calls, we are fairly confident that Adobe can win in the professional space despite the rise of AI.

Before going over Adobe’s competitive advantages in the professional space, we should mention two things to give some context. Firstly, Adobe has already released a similar tool to Dall-E and Midjourney called Firefly. Secondly, while generative AI can create “direct” output if you are just looking to have fun, companies that want to use this content commercially will likely have to make some editing to the output. This should make generative AI a complement more than a substitute for current design tools.

So, how does Adobe differentiate in the professional space?

Adobe Competitive Advantage #1:The Data To Train The Model

Generative AI models require a lot of training data. Gathering this data is extremely expensive, creating some entry barriers. According to Adobe’s CEO, Shantanu Narayen, there are only a handful of companies with the data-gathering capacity to come up with leading-edge AI models (emphasis added):

The fundamental models, whether it be a text model or a visual model, there are 3 or 4 companies on the planet that actually have the capability and the science to do that.

Source: Adobe Summit Q&A Session.

This makes sense on paper, but if this is really the case, why are we seeing so many upstarts come up with generative AI models? There might be two explanations for this. First, it’s a hot space right now, and there’s probably a lot of VC (Venture Capital) funding flowing into AI. µ

This said, the most likely explanation is that these upstarts have followed the strategy of building a community that trains their models for them with questionable data (more on this later). When you converse with Open AI’s ChatGPT, you generate data that is subsequently used as input into the model to train it further. More interactions create more training data, eventually making the chat better and more appealing to more users. This powers a flywheel.

Something similar happens in generative AI, where the images that come out as output are eventually fed into the model to develop it further. This flywheel weakens the monetary entry barrier, but the data quality also matters. Adobe has millions of users creating high-quality digital content daily on their apps. I would even argue no other company can equal the quality of this content library. The quality and the sheer size of this data can help Adobe end up with a better generative model than competitors, but it’s not only quality and size that matters…the origin also does.

Adobe Competitive Advantage #2: Commercially Safe Content

The other day, while scrolling on Twitter, we came across the following tweet, which compares Adobe Firefly to Midjourney, a self-funded 11 people team:

While one would think that Adobe’s Firefly should have produced the image on the right, that’s Midjourney’s. The difference between both images is a question of the data both companies used to train the model.

Adobe only uses non-licensed material, whereas almost every other model uses both licensed and non-licensed material. As per Shantanu Narayen (emphasis added):

A lot of the other companies are actually using data that potentially…they’re scrapping the Internet, they’re accessing data that they may or may not have rights and license to.

Source: Adobe Summit Q&A Session.

This means that if you ask Midjourney for a famous character, then the AI will simply show you that character, whereas if you ask Firefly, the AI will not show you that character and will try to come up with something similar. If you are asking for Mario or Pikachu, the best outcome is to see just that, but in comes the concept of “commercially-safe content.”

Midjourney’s output is not commercially-safe, and professionals would most likely refrain from using it. In this case, using Mario or Pikachu would require explicit permission from Nintendo (owner of those characters), and if not given, then the user of those images could face litigation and reputational issues, which are already ongoing today (emphasis added):

I mean you’re going to see a whole bunch of other companies, they’re open source. And the question is what data went into it? And clearly, there’s a fair amount of litigation also going on.

Source: Shantanu Narayen, Adobe’s CEO, during the Adobe Summit Q&A Session.

How many corporations are willing to risk using a generative AI model that can lead to reputational or legal issues? I’d say few. Adobe knows that many of its customers fall in this segment, and thus has been very careful to build Firefly as a commercially-safe generative AI model (emphasis added):

There will be other models that exist. And certainly, there’s the issue of the more data that the model has, the better it is. The hard problem is doing it the way Adobe did, which is doing it in a commercially safe way.

Source: Shantanu Narayen, Adobe’s CEO, during the Adobe Summit Q&A Session.

One of the problems of closing the data gap by making the model open source is that the input to train the model can carry licensed elements, making the output carry these flaws too. We don’t know if there’s a way to control the input or to filter it so the output is not affected, but our feeling is that the origin of the data you use to train the model matters quite a bit.

So, not only do companies need large amounts of high-quality data to train their generative models, but they also need this data to be commercially safe for use by corporations. We have a hard time seeing how Adobe does not have an edge here.

On top of that, to Adobe’s defense, the product is in an earlier stage than that of Midjourney, but that can change fast.

Adobe Competitive Advantage #3: Product Breadth

Before discussing Adobe’s competitive advantages, we mentioned that professionals would most likely need to edit the output of a generative AI model. While these models are pretty accurate, it’s highly unlikely that they “spit out” exactly what the designer is thinking of with just one text prompt. This is where Adobe’s Creative Cloud comes in handy.

Many generative AI programs are single-point solutions, and thus, users would most likely have to use other programs to edit their output. The problem is that if a designer inserts a photo made with a program that’s not Firefly into Photoshop, they will not be able to identify the different vectors (or segments on the photo). Let’s clarify this with an example management gave in the most recent Summit.

Suppose a designer asks Firefly to produce an image of a colorful bird. After Firefly produces the image, the designer wants to edit the wings in Photoshop. As the image was created with Firefly, Photoshop will automatically identify where the wings are. If, on the contrary, the designer has come up with this image using Midjourney and inserts it in Photoshop, it will not be able to identify the different vectors, or at least accurately (emphasis added):

And you can say, generate me a picture of a bird. And it will generate you a picture of the bird and you downloaded it. It’s probably downloaded as an image or SVG. If you want to use that in Illustrator to create some artwork associated with that, you have to vectorize that. Trying to vectorize an SVG image, sure, it will vectorize it. It will give you 40,000 spots.

Then try and edit that. It’s impossible if you have done it from somebody else. If you do it in Illustrator and you do it with our technology, it’d be like, I know this is a vector.

Source: Shantanu Narayen, Adobe’s CEO, during the Adobe Summit Q&A Session

This seamless connection between generative AI images and the Creative Cloud is Adobe’s way of adding value through a comprehensive solution.

Adobe Competitive Advantage #4: The “Content Supply Chain”

For most companies, digital content creation is just one step of the entire process. These companies typically use this content as part of a marketing campaign, so they are interested in understanding how the content performs and how they can improve these campaigns. Adobe refers to this as “the content supply chain,” and it’s where the Experience Cloud comes in.

Adobe started the Experience Cloud some years ago, not as an independent business line but as an add-on to the company’s Creative Cloud. The goal was to provide customers with a single place to create digital content and measure its performance. There’s no other company out there that provides this, which constitutes a strong competitive advantage for Adobe, especially in the enterprise space.

Of course, companies can definitely create digital content with a non-Adobe program and just purchase the Experience Cloud independently, but large corporations tend toward simplification, not complexity. This makes the connection between Experience and Creative Cloud a compelling value proposition.

So, Adobe envisions the creative workflow with generative AI as:

1. Creating an image with generative AI to avoid working from a blank piece of paper

2. Introducing this AI-generated image in Creative Cloud apps to make some edits

3. Measuring the performance of the campaign where the image is used through the Experience Cloud

4. Iterating on the campaign in real-time and repeating the process.

There’s no company out there today that can replicate this workflow with proprietary tools, which plays to Adobe’s strength in the enterprise space.

Conclusion

Adobe Inc.’s generative AI technology might seem worse than that of competitors, but the company has four strong competitive advantages that will allow it to continue being the best value proposition in the enterprise space. To this, we have to add that while Adobe’s Firefly is evolving with commercially safe content, other competitors can’t say the same and will probably face litigation issues down the road.

How generative AI will impact the non-professional market is another question. We think Adobe will struggle in this segment, as non-professionals don’t care too much about commercially-safe content or having an end-to-end process with the Experience Cloud. Will regulation advance to the point where generative AI models trained with licensed content will be banned entirely? Time will tell, but knowing Adobe is not at risk of such litigation headwinds is a plus.

We rate Adobe as a “Hold” until the uncertainties surrounding the Figma acquisition are dissipated.

In the meantime, keep growing!

---

Analyst’s Disclosure: I/we have a beneficial long position in the shares of NTDOY, ADBE either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

---

Best Anchor Stocks helps you find the best growth stocks to outperform the market with the lowest volatility. They can anchor your portfolio on the stormy market sea, still allowing you to outperform.

Best Anchor Stocks have a long track record of revenue growth combined with below-average volatility. The first pick is up 16,000% from its IPO but has never been down 30%, not even during in 2008-2009. 

Best Anchor Stocks can serve several purposes: stabilize your high-growth portfolio, or add low-volatility growth to your index, dividend or value investing.

There’s a 2-week free trial, so don’t hesitate to join Best Anchor Stocks now!