Adobe: Is OpenAI’s Sora Changing The Game? (NASDAQ:ADBE)

Introduction

OpenAI launched its new text-to-video model “Sora” a couple of weeks ago. The market did not like what this announcement supposedly meant for Adobe Inc. (NASDAQ:ADBE) and sent the stock down significantly. The stock was down around 13% in a matter of 5 days:

This article brings my thoughts on this topic, with the goal of answering the following question:

Is OpenAI a threat to Adobe?

Note that investing will always be subjective because we need to envision how the future will look like. This means that scenarios tend to be grey rather than black or white, something that will not be different this time.

I’ll explain what OpenAI’s “Sora” is, the potential problems for Adobe, together with “the other side of the coin.”

Without further ado, let’s get on with it.

What is OpenAI’s “Sora”?

Sora is OpenAI’s new text-to-video AI generator. Sora allows its users to generate videos based on a text prompt. In essence, it works very similarly to Dalle-E or Firefly (text-to-image generators), but it’s more impressive because video is a dynamic format that includes interactions and 3D generations. For anyone who has seen the clips, there’s no denying that the results of Sora seem extremely impressive.

Sora is the next step of the so-called “generative AI” trend, enabling people without prior experience to create digital content.

The potential problems for Adobe

The launch of OpenAI “Sora” has worried the market about what it means for several companies like movie studios, ad agencies, and editing tools, among others. Adobe belongs to this last group of companies, and I will summarize what I see as potential problems for the company in this section.

The first and probably most obvious potential problem is that the future now looks more uncertain than it did several years ago. AI is breaking the barriers to content creation, and we can’t grasp its limits. One can argue for or against its commoditization, but one can’t deny that we don’t know what’s in store for us in the future. Will Adobe remain a leader in the creative industry 10 years from now? Regardless of the answer to this question, I’d say it’s more challenging to answer it confidently today than it was before the arrival of gen AI. Current AI developments are primarily impacting the generative part of the process, but we are already seeing some hints that it might also impact the editing phase from which Adobe generates most of its revenue. Adobe’s generative fill feature is a good example of these early hints.

This said, as we must focus on the future to invest, we must assess what this future will look like. The first implication for Adobe is that the number of seats might be negatively impacted by the increases in productivity brought by AI. The reasoning is straightforward: more productivity means fewer seats needed to conduct the same tasks. This would obviously be negative for Adobe as its subscription revenue depends on the number of people using its creative suite. There’s a caveat, though.

Adobe’s management has stated that they aim to monetize AI through a per-use model in which users spend “x” credits every time they use Adobe’s gen AI capabilities. AI might reduce the number of seats, but part of this “downfall” would be netted out with higher revenue from generative AI. Judging whether this will be a net benefit or a net negative is still out of my reach, especially since this new monetization model is new and we don’t have much data about its success.

The other potential problem I see comes from distribution. When AI became a buzzword, many claimed it would completely disrupt all software companies. While I understood these worries, I never thought AI would be the disruptive force many claimed it would be because distribution matters dearly.

If a company builds the best AI model but has no distribution, the player with the distribution will likely be able to commoditize it fast enough so that it does not gain widespread adoption. This is why I don’t worry too much about AI startups competing against the “big guys.” This said, OpenAI has a relationship with probably the largest “guy” of all in enterprise software: Microsoft Corporation (MSFT). This makes OpenAI not just an AI startup, but an AI startup with access to a significant distribution channel.

There’s no denying the future looks more uncertain today than it did a couple of years ago. AI might pressure the number of seats available in the industry, and OpenAI should not be discounted as an “AI startup with no distribution”; it’s definitely a potential threat.

The other side of the coin

As in all things investment-related, there’s another side to the coin. Arguments are rarely one-sided in investing, and this one is no different. There are also reasons to believe Adobe can continue to thrive in the future and that this “AI wave” can ultimately benefit the company.

The first one is that generative AI is unlikely to provide the end result the creative seeks. Sora (or any other generative AI model) might provide the creative with a video based on a text prompt, but it’s improbable that the video will be exactly what the creative is looking for. This disparity (between what the creative wants and what they get) might be acceptable in the non-professional world, but I highly doubt professionals will take this output as their end result.

This gap or disparity means the creative suite of tools will probably remain relevant as generative AI would only “solve” the first step. This has already happened in AI-generated images, and it’s tough to envision a scenario where it doesn’t happen for video. Note that this is especially true considering that video is a much richer format with many more potential divergence points. This begs the question…

Why would AI-generated video be different than AI-generated images?

When OpenAI released Dall-E, everyone thought it would completely disrupt Adobe. However, the company replied by commoditizing it with the launch of Firefly, which it embedded in its creative suite. With management repeatedly stating that video would be the next format to which generative AI would be applied, why should this time around be different? If anything, AI-generated images have been beneficial to Adobe (to date).

Sora might just get “commoditized” as Dall-E was, and we should not forget that Adobe owns the entire value chain from generation to editing to measurement (through the Experience Cloud). This whole value chain is very relevant for professionals, and customer groups where most of the monetization opportunity lies.

Generative AI is making the creative industry more accessible by lowering the entry barriers. This has its pros and cons. The main con is that switching costs are lowered so the moat might contract somewhat. The main pro is that the potential pool of seats might expand significantly. The only caveat here is that much of this pool might not be monetizable, so the thesis relies on how Adobe manages to price to value its AI offerings to the pool of monetizable users.

The good news for the company is that these users care more about the entire value chain than just the generative phase. There are clear synergies to having all phases under the same umbrella.

Another point worth mentioning is that Adobe believes that despite AI bringing more productivity, it might expand the pool of users as companies look to engage in much more content generation. This was well encapsulated by Adobe’s Chief Product Officer, Scott Belsky (emphasis added):

Now this is consistent with job growth in previous platform shifts, whether it be engineers becoming increasingly more efficient over the last few decades, the advent of no-code web builders in the early 2000s, or the rise of social and user-generated videos over the last 2 decades. Through all of these changes, the need for more engineers, more web developers, and video pros, these needs only grow.

And the results actually from our own customer surveys are also encouraging. We’re starting to really connect with our customers. We’re trying to ask them what all this means for them, what they’re perceiving.

Why I rated Adobe as a hold

Many of you might already know that I have a clear preference for not selling unless I have a clear view that the thesis has been permanently impaired. This bias stems from the asymmetric nature of stock investing: stocks have (virtually) unlimited upside, but the downside is capped at 100%. This means that the most costly mistakes tend to come from selling a winner too soon and missing on all of the upside rather than selling a loser too late and suffering a 20-30% additional drop.

Don’t get me wrong, I definitely care about the downside, and it’s one of the most important considerations in my research process. Still, I also think that selling a great company when it’s unclear if the business is deteriorating might turn out to be a mistake.

Could the wave of AI end up deteriorating Adobe’s business? The answer to this question is unclear to me currently, and if Dall-E serves as an antecedent, I think we tend to overestimate the threats. I will definitely sell my position if I get convinced that generative AI will deteriorate Adobe’s business. I don’t mind doing this at a 20% or 30% lower stock price; this is the price I am willing to pay to avoid missing on the upside.

I already trimmed Adobe once back when it announced the acquisition of Figma. That turned out to be a costly mistake because the stock has gone up significantly since my sale. Then came Dall-E and I thought I made a good decision trimming, but Adobe quickly launched Firefly and retained the entire value chain. It takes something substantial to turn a high-quality business from great to bad, so I am willing to hold my position and wait until I see strong evidence of such deterioration.

Additionally, I would like to wait for Adobe’s earnings to see what management has to say on this topic and get more clarity on what the company is doing in text-to-video generative AI. Adobe will report earnings in several weeks, post-market on March 14th.

Conclusion

I hope this article helped you understand my views on OpenAI’s Sora and what it means for Adobe. I have a bias towards not selling unless it’s obvious, and this situation definitely does not seem obvious to me.

In the meantime, keep growing!