1/ My first deep dive is live. Won’t be my last.
I spent a month learning biology to understand one company: $NAUT.
They built a machine that reads proteins one molecule at a time. Nobody else can. 🧵
Speech to speech is the biggest tailwind $DUOL has ever had.
Speaking is what people actually want to learn, and it's always been the weakest part of any language app.
Voice AI can now hold a real conversation. You talk, it interrupts you, corrects your accent and adapts in real time, the way a good tutor would.
Anyone with an LLM can build a voice tutor. Very few people would stick with it.
The hard part of learning a language was never the teaching. It's showing up every day for months.
$DUOL has spent over a decade learning what makes people come back. Streaks, leagues, notifications that land at the right time, the owl guilt-tripping you. Millions of experiments on human behavior that a new app would have to start from zero on.
And it's social. Your friends, your league, the streak you share with someone. A chatbot can't give you that.
Better voice AI makes every one of those sessions more valuable. The habit is already there.
Learning is about to get more addictive than ever.
NFA - $DUOL is one of my largest positions.
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
https://t.co/r0qFLdvd6c
My bet is we will see big reacceleration in 2027 and 2028 of user growth and paid subs. Target 35m paying subs by end of 2028, $4bn revenue, 40% FCF margin, $1.6bn FCF at a P/FCF of 30x. This equates to roughly $1000 per share by end of 2028. That’s probably the high end. Low end $800.
Post 2028 this will be the dominant global education company gearing towards 1bn global learners and billions in FCF. That’s my ultimate thesis here on a 5-10 year horizon as AI exponentially improves learning outcomes.
Long term hold
@MichaelSagun777 167. I liked it in the 200s and I’ve kept buying on the way down. I keep buying still. It’s pretty much all I think about, how to buy more and more 🤣
@BiologyBull Maybe we should look into buying some Voyagers then. Especially if the payback period is less than 2 years. Sounds familiar!
And I’m not joking lol
Picture robots running assays on Nautilus around the clock, pushing more and more samples through the system. Mapping proteoforms will never be finished: there are likely millions of them, they change with age, disease and drugs, and each new target is another assay kit. If the thesis plays out as we expect recurring revenue should rise, and translate into more and more FCF
They'll use whatever is cheapest and gives members the best learning outcomes.
I know they've been moving most of their AI workloads to open-source models, which has drastically cut their per-unit AI costs.
As for this tech from Jack, I'm sure it's just the beginning. Eventually it will be pretty commoditized, and the value will go to $DUOL's algorithms that keep you engaged, not to whoever owns the voice tech. Learning will be more addictive than ever.
Congrats on the streak!
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
https://t.co/r0qFLdvd6c
@BiologyBull yes as the installed base grows, the recurring revenue will grow with it. $500k will eventually be much higher as the utilization becomes pretty much 24/7.
$DUOL will be better than any tutor. For this to happen, speech to speech for video calls and speech recognition exercises is crucial. It will be able to teach you real time and correct you, and further personalize the teaching experience.
When this becomes apparent the investor community will wake up and realize $DUOL has no competition when it comes to education.
Random thought: once AI glasses look indistinguishable from regular reading glasses, closed book exams effectively stop working. The future of education is going to change whether schools are ready for it or not.
Skill of the future is presenting ideas coherently out loud.
Reminds me, I will start doing more video format…coming soon.