Why AI Has Me Bearish on Adobe, But Not Duolingo

Many businesses are being sold off because of AI fears. Two prominent SaaS companies affected are Adobe and Duolingo.

Adobe has crashed by 43% since February 2024.

Duolingo has crashed even harder. Down 65% since May of this year.

Is AI really going to wreck these companies? Here’s the bear case being presented:

Adobe’s AI Bear Case

Adobe has been the de facto standard for audio and visual content creation for over 30 years. While it built one of history's most successful SaaS empires, bears fear that dominance is finally ending. The primary reasons are:

• Competitors are moving upmarket. Canva and modern rivals are aggressively targeting enterprise customers. Their ease of use, combined with AI, allows non-professionals to create content without navigating Adobe's complex legacy software.

• Seat compression. AI makes users drastically more efficient. Enterprises will simply need fewer software licenses to produce the same marketing output.

• Generation replacing editing. Tools like OpenAI’s Sora and Midjourney are bypassing the need for traditional editing entirely. Investors fear a future where content is generated rather than edited, making tools like Photoshop or Premiere Pro less essential.

To visualize just how fundamentally unchanged Adobe has remained over the last three decades, take a look at these screenshots.

Original Adobe Photoshop Circa 1990

Recent Adobe Photoshop Circa 2024

In the 30 years since launch, Photoshop is the same. It’s an image editor where you create or edit an image and use a bevy of standard tools to make changes.

Virtually every Adobe software is the same as their original release, although equipped with newer tools layered on top.

If we look at software like Adobe Premiere Pro for video editing, it’s also become outdated to many offerings of modern peers.

For example, in Descript, a competitor for video editing, you can do all of the following:

• Instantly Remove Retakes: The AI scans your recording, identifies where you repeated a line, and lets you delete the bad takes with one click, leaving only the best version.

• Overdub AI Voice Cloning: If you misspoke a word (e.g., said "Monday" instead of "Tuesday"), you can just type the correct word. Descript uses an AI clone of your voice to generate the new audio and seamlessly patch it in, no re-recording needed.

• Auto-correct Eye Contact: An AI effect that subtly adjusts your eyes in the video to look directly at the camera, even if you were reading a script off-screen.

• AI Generated Viral Clips: An integrated AI that can auto-generate viral shorts from long-form content, write social media captions, and summarize your video.

While Adobe integrates some AI features and users can find 3rd party integrations to offer some of these capabilities, many AI features are not native. New competitors like Descript, with a more modern platform, have them fully integrated, which makes them easier to use and more capable.

These new features often make obsolete the cumbersome editing processes that the original Adobe suite was designed for.

Adobe’s product has always been giving people the ability to create and edit things. Now the way we edit things in 2025 is changing. We may no longer need the same visual framework of Photoshop from 1990, or Adobe Premiere Pro of 2003.

Why I Am Bearish on Adobe

I’m bearish because the fundamental products Adobe offers are not universally the best product anymore. They were designed to be the best product when they were released in the 90s and early 2000s.

If you want the best image editing software or best video editing software today, that is, software that saves you time with fantastic finished products, you would not design it to work like Adobe.

Adobe tools are complex, cumbersome, and slow to use compared to modern competitors. They offer some advantages in specific technical use cases, but as competitors continue to evolve, the technical benefits shrink.

People are not using competitors like Canva or Descript to save money. They are using them because the product itself is better for their use cases.

As the need for the “legacy” visual organization of Adobe products declines, so too will Adobe’s market share in the space.

The speed and rate of this decline may vary. Legacy users could be very sticky, with Canva and others continuing to take new cohorts. Or, generative AI could become so successful that it entirely removes the need for Adobe’s legacy visual interface.

For these reasons, I am bearish on Adobe, and uninterested in the “dip”, which I view as risk-off for a mature SaaS company showing stagnant growth, and potentially entering decline due to technology shifts.

Duolingo’s AI Bear Case

Like Adobe is the leader in its industry, Duolingo is also the de facto leader in online education, focusing primarily on language learning. It’s maintained its position as the #1 Education app in the app store for eight years straight.

Bears also fear that Duolingo’s top position could be maturing. Let’s take a look at the bear case:

• Generative AI creates superior, personalized tutors. While Duolingo relies on structured, gamified drills, general-purpose LLMs now offer infinite, open-ended conversational practice and instant grammatical explanations. Bears fear users will migrate to these "smart" tutors that adapt to specific needs rather than following Duolingo’s rigid, linear path.

• Real-time translation erodes the core purpose. As technology allows for seamless, real-time translation the functional need to spend hundreds of hours learning a language diminishes.

• New competitors can launch easily with AI

So, given the above, why am I bullish on Duolingo, while I am bearish on Adobe? Aren’t they both being disrupted by AI?

First we have to understand what the product actually solves.

Adobe - Is generally used to create a finished product, often marketing related, typically used for business purposes. The user generally does not use Adobe because they “like Adobe”. It is merely a means to an end. If there is a quicker and more efficient “end”, they will use that instead.

Duolingo - Is an app that is “fun” to use. It delivers dopamine to users through gamification of learning. Many users say if they didn’t use Duolingo, they would be scrolling social media instead. The reason Duolingo users may initially download the app is for “education” or a desire to learn something. The reason users stay is for the dopamine, the psychology/peer pressure they implement, and the general feeling that they are getting something out of using the app.

The key difference between the two is Adobe is a means to an end. Whereas Duolingo is about the journey.

Given this, here is my rebuttal to each bear case.

Generative AI creates superior, personalized tutors?

The assumption here is that Duolingo is primarily providing benefit through tutoring. The problem with education and learning is primarily retention. Most people have an initial urge to do something, such as learn a language, play the piano, or go to the gym.

The problem is retention. People easily lose interest. Hence, gym use is always highest in January, but invariably peters off as the year progresses.

Anyone can use Gemini to practice Spanish. But, after a few days, the majority of people without strong intrinsic motivation get bored and move on.

This is the actual problem Duolingo solves for. They create the best system for retention in education. In fact, there are millions of people with a 365 day streak on Duolingo. There are even people with 10 year streaks on Duolingo.

Once you use Duolingo you are onboarded into a dopamine peer-pressure notification system that gets people back day in and day out.

That brings us to the next bear case.

Can’t other people do this too with their own SaaS software thanks to AI?

Yes, anyone can make a Duolingo app clone. Or a tutor they believe is better. But how do they tackle retention?

Duolingo has two huge advantages here:

1. Duolingo generates 1 billion data points on retention every day

2. Duolingo has a network effect

Duolingo benefits from the network effect by being the #1 education app for 10 years. Rosetta Stone never had a social aspect like that. Duolingo encourages you to share learning streaks with your friends. While this sounds simple, it creates strong peer pressure. Not only would you break your own streak for not learning, you would break your combined streak with a friend. Duolingo can do this because they are the de facto language learning and education software where everyone goes to. And people will go where their friends are.

More important than this, Duolingo generates over 1 billion points on retention each day. If you are a competitor, you have absolutely no chance of matching retention. Duolingo knows when to send you a message, what type of message will make you most likely to use the app again, how intense to make each learning session, etc… Their A/B testing in the Duolingo system has no parallel.

The final problem for new competitors is that Duolingo is willing to make new features. That means if you are a new competitor who somehow manages to come up with a new way to teach language, Duolingo can simply copy your method and make it better with their massive user base to A/B test for them. Just like Meta/Instagram copied Snap’s features, Duolingo can copy competitors’ lesson structure.

Adobe, unlike Duolingo, doesn’t benefit from generating A/B retention data from customers to improve the product. It also doesn’t benefit from the network effect.

At the end of the day, all Duolingo competitors are still stuck to the same system as Duolingo, they use the smartphone and an app. They have to play by the same rules as Duolingo. But they don’t have the A/B testing capabilities or the network.

What about native language translation?

As technology improves, there will always be the desire to learn. Unlike Adobe, Duolingo is about the journey, not the destination.

People want to learn languages because it’s fun. Because they want to understand people themselves when they talk, because they legitimately enjoy learning, and because the app is fun and delivers dopamine.

Live translation is solving a different problem. In live translation, someone is actively communicating to you and you need to understand what they say. This kind of software serves a practical immediate need.

Duolingo doesn’t have millions of users in the US learning a language everyday because they need a translation of a business meeting for their job. They are on the bus and want to do something fun and learn at the same time, so they open up Duolingo. If Duolingo didn’t exist, they would open up Instagram instead.

Why is Duolingo being sold off then?

Duolingo is being sold off because of growth deceleration in mature markets. Let’s take a look.

There is a steady decline in the rate of growth for Monthly and Daily users, and Paid Subscribers.

Not only is the rate of growth declining, it is all in relative unison.

If we track this decline vs share price, we will see that Duolingo has had 5 quarters of significant growth decline, beginning in September of 2024.

Strangely, the share price actually went up dramatically in early 2025, despite paid subscribers continuing to fall.

Perhaps investors were forgiving in this area, as DAU remained relatively strong until a precipitous drop of 8% points (in terms of growth) in March 2025.

Adobe vs Duolingo From a Valuation Perspective

So which is the better business in terms of valuation?

Let’s compare each business from a practical valuation perspective, starting with EV/Free Cash Flow.

Currently, Duolingo is about 38% more expensive on a EV/FCF basis.

So, how much more growth are you getting for that premium? See below.

As we see above, Duolingo has grown FCF triple digit over the past few years. Adobe, by comparison, has been flat. It’s fairly obvious to see that if Duolingo can continue to grow for just a few more quarters at its current rate, with share prices remaining flat, its valuation will already be matched with Adobe.

However, once we arrive at that point, Duolingo will still be growing faster than Adobe, and for reasons I previously mentioned, will be less likely to incur continued declines due to the threat that AI poses to the SaaS industry.

For these reasons, I much prefer Duolingo at today’s share prices than Adobe. You are ultimately getting a faster growing company at what I consider a more reasonable valuation (noting that stock based compensation is still significant for Duolingo).

While Duolingo does have a declining rate of growth, it is natural as SaaS companies grow into their TAM. Duolingo may be able to increase that TAM through new offerings, though current ones haven’t significantly shifted this trend.

My base assumption is the rate of growth will continue to decelerate; however, Duolingo’s superpower is its retention. New cohorts will age in every day, and as world economies grow and prosper, Duolingo will also benefit.

New offerings are also young, and chess could prove a larger social network for gaming, which evolves into a variety of games. Player vs Player capabilities in Duolingo only launched for Chess a month ago. This new suite of products could buoy users more than we expect.

What do you think? Will AI help or hurt Adobe or Duolingo? Which stock is the better pick going into 2026? Leave a comment below.

If you enjoyed this article, you may also be interested in reading Jim Cramer's 10 Failed Dot Com Bubble Stocks & Why They Failed.

ALSO READ: The Generational Shift in Distribution That Will Define the Next Decade of Winners

Our content is provided subject to our Disclaimer.