🟢 BehavioralEdge 🟢 (@_MarketBehavior) header
🟢 BehavioralEdge 🟢 (@_MarketBehavior) profile picture

Investing in social behaviour: predictive and disruption-resistant. Exploring the human code. $HIMS $DUOL $META $GOOG $AMZN

Mediterranean Europe
Joined March 2025
I just realized a key issue in education that explains $DUOL ’s past, present, and future trajectory. Educational outcomes do not primarily correlate with teaching efficacy. The two main factors that explain teaching outcomes are: - Alumni selection. - A physical ecosystem with high talent density. The only exception is one-to-one tutoring. Luis and Severin knew they could not compete in selection and ecosystem, but they could aim for the only exception. We can define two distinct eras: Pre-Q3 2025 and Post-Q3 2025. The pre-era was about building a world-class machine, ready for what was about to come: the moment when one-to-one tutoring, powered by AI, could finally be made universally available. The post-era is the first time in the company’s history that it is focusing entirely on teaching efficacy. Q3 2025: "In Q3 we decided to shift the balance towards longer-term initiatives. In particular, we’re investing proportionally more in teaching better." Q4 2025: "Here’s our plan: Teach languages better." Q1 2026: "So we are prioritizing teaching better."
1
2
2
26
6,774
Huberman has just started to pitch $HIMS
Peptides are not as safe as you think: People assume that because peptides aren't hormone therapies like testosterone or estrogen they are innocuous and free of side effects. That assumption is wrong. These are very potent compounds with many different effects and we are in the early stages of exploring them. If you use them: work with a very good board certified physician, make sure the sourcing is clean, do regular blood testing and monitor for any potential tumor growth. @hubermanlab
2
3
1
125
10,263
$HIMS When people start researching peptides, they are highly likely to come across Andrew Huberman. He combines two extraordinary assets: enormous public reach and an exceptional scientific reputation. There are plenty of Serena Williamses, Megan Foxes and Charles Barkleys. But there is only one Andrew Huberman, and $HIMS is the one who got him onboard.
2
2
38
2,151
🟢 BehavioralEdge 🟢 retweeted
I’d like to formally invite $HIMS CEO @AndrewDudum to the pod Would be great to have @hubermanlab on simultaneously I think retail investors are not understanding the strategy $HIMS is executing and this pod would help
76
95
4
1,394
59,491
Added $HIMS twice today, plus some very OTM Dec '27 calls. Probably one of the things that makes people more emotional is deceleration. Because it is quantitative. Children aged five to seven already know how to recognise patterns in sequences of numbers. Our brains will instinctively construct a narrative as soon as we recognise a pattern. But this pattern, in fact, provides poor information about the underlying health of the business. Focus on qualitatives. Not Financial Advice
2
1
10
981
$HIMS Hubermann cannot afford being perceived as having compromised scientific integrity for money. Scientific reputation is fragile. It has taken him decades to built it. His appointment is strong evidence of his confidence in the scientific and ethical credibility of $HIM's direction.
2
2
55
3,297
$DUOL just have to wait and benefit from the agent race. "enables to actually catch these things like interruptions and simultaneous speech and really create a rich, high bandwidth interaction between humans and machines, which we think is critical to actually enable agency"
This is f*cking insane. Former OpenAI CTO, Mira Murati: "The models we work with today are turn-based. You talk, they talk, then they go off and think, and while they're thinking they're deaf and blind. The next ones never stop listening. They take in audio, text and video all the time and answer in 20-millisecond chunks." In just 2 minutes, she shows why the chatbot you type into is about to look like a fax machine. Prompt Box → Chatbot → Agents → AI That Works Next To You In Real Time A turn-based model waits for you to finish typing, then goes quiet while it thinks. An interaction model sees and hears what you're doing, and can cut in the moment you start going the wrong way. Anyone can type a prompt. Almost nobody is ready to work with an AI that never stops paying attention Watch the full 2 minutes. This is worth your time. Bookmark it, like it, save it, I don't care... this is the best thing you'll see on X today.
1
7
1,039
$DUOL Brand emotional attachment. DUOMART OSAKA November 7, 2026, runs until November 17. The previous DUOMART, held in Tokyo in December 2025, attracted approximately 10,000 visitors over 13 days. $DUOL is developing something that Disney, Nintendo and other entertainment brands have successfully exploited: turning digital characters into physical experiences, merchandise and emotional attachments.
デュオマート大阪も初日から大盛況みたい☺️ 中国のイベントもよくやってるイメージだしアジア市場はかなり重要なんだと思われる。
1
5
349
🟢 BehavioralEdge 🟢 retweeted
As a pharmacy owner, $HIMS could put me out of business whenever it wants. Here’s why I’m bullish on the stock. Getting testosterone cypionate through my pharmacy: Doctor’s consultation: £300 Blood tests: £350 Testosterone cypionate (3-month supply): £233 Total: £883 (~$1,185) for 3 months Getting it through $HIMS: $79/month = $237 for 3 months. Same treatment. A fraction of the cost. This is what disrupting healthcare looks like.
3
17
1
140
6,022
Me yesterday, after $HIMS' strong move: rinse and repeat. Shares purchased at $23.81. Jan '29 LEAPS Puts sold at a $45 strike, collecting $25.00 in premium. I cannot leave that money on the table when there is such a significant disconnect between perceived and actual risk. Moreover, in the AI era, faster iteration cycles have made LEAPS strategies substantially less risky. Not financial advice.
I just sold $HIMS Dec 2028 puts and bought more stock. Buy price–value dislocations. Sell volatility when perceived risk massively exceeds fundamental risk. From first principles, investing is about exploiting dislocations: price vs. value, and perceived risk vs. fundamental risk. Not financial advice.
1
3
430
I first bought $HIMS in November 2024 at $24.70. Two years of flat share prices. Fundamentals relentlessly moving up and to the right. Thank you for the drawdown!
People deserve to have access to the most innovative treatments. That’s why we work with leaders across the industry, and starting today, that includes @EliLillyandCo. When innovators work together, it’s better for everyone. Customers can access more personal care, providers have a well-rounded picture of how their patients are doing, and drugmakers can be confident that the people who need their medications the most will be able to get their hands on it and stick to it. Lilly is one of the most innovative companies in the world, and we share a commitment to reaching real people with real impact. Working together means we can combine some of the world’s best medicines with our powerful platform to help more people feel great. Read more, including important info: news.hims.com/newsroom/lilly…
1
2
32
1,186
Thank you for the drawdowns!
Your depression is my opportunity
3
277
My average $DUOL price after options premium expected deduction: $95-$105 I rolled again Jan '28 Puts to Jan '29. Cashed some gains, decreased risk. Thank you so much for the drawdown!
End of last week, I sold $DUOL Jan 2029 Puts and bought more stock. All the biggest 9 language learning apps are broken businesses, except for a world-class business. All offer lifetime subscriptions, except $DUOL. AI agents will do work for you. AI tutors won't. No AI lab will enter this garden, because there are no incentives. I don't see any clouds in the horizon, except the challenge of an extraordinarily complex mission. Not Financial Advice
1
9
668
Congrats $DUOL shareholders! 3Q 2026: 2028 targets: on track. Good news, but nothing highly relevant. I'm working on a 1 Trillion Roadmap. I have identified multiple qualitative/quatitative KPIs to follow. Luis talked about 1: Personalised Tutoring Adapted to Cognitive and Emotional States "I mean, by the way, the reason one-on-one human tutors are better than a classroom or better than what we do right now is that they are really good at adapting to you. They're just -- and they do even things like they read your facial expression. When they're saying something to you, they can see that you're, like, you look confused. So you don't even have to say and then they quickly adapt to you. That's why they're so good. And this -- I believe the technology to do that is just getting better and better, and we'll be able to do that. It'll take some time, but we'll get there, I think." My Deep Dive: Personalised Tutoring Adapted to Cognitive and Emotional States At some point in the future, we want to hear Duolingo’s management explain how the company will adapt its teaching methods in real time to each student’s cognitive and emotional state. Duolingo’s existing voice calls and future AI tutors will interact with learners through spoken conversation. On October 23, 2026, Jack Dorsey’s company, Block, announced Dessa, an AI voice model that it claims can detect emotional cues directly from speech, rather than relying solely on the words being spoken. Wearable devices can already monitor physiological indicators to deliver personalised recommendations. The same approach could be applied to education. Whoop, for example, already uses metrics such as heart rate variability, resting heart rate and sleep quality to provide personalised exercise recommendations.
1
18
784
🟢 BehavioralEdge 🟢 retweeted
Some thoughts on compounding free cash flow per share in the AI race, after this week’s $PLTR, $SPOT and $META earnings 1/ There are now two distinct ways to do it. Own the data: build infrastructure that generates proprietary data, then post-train any available model to deliver more value per token than anyone else. Or own the frontier: push the frontier of intelligence and sell it at a premium 2/ $PLTR’s Q3 shows the first can disrupt the second. Its customers are beating frontier models on their own tasks by post-training open-weight models on proprietary data. At a PACOM exercise, a 30B model beat a 3T frontier model. L3Harris beat the frontier model it had been using in under 48 hours, with costs down 95% in preliminary results. McCarthy got better-than-frontier results from a 4B Nemotron and will save 30,000 hours a year 3/ For those who own the frontier, this is deflationary: every cheaper model that clears the bar erodes their premium. For those who own the data, it’s the opposite. Their cost of intelligence falls while the value of their data rises 4/ $SPOT is the hallmark of running this playbook inside a P&L. It routes each task to the model that fits, open-source included, and spends those cheaper tokens on personalization no one else can generate: custom mixes, personal podcasts, book recaps. Every mix and agent request feeds its taste model, which is already lifting discovery by double digits. The result: 100M+ users active 28 days a month, ARPU +7.5%, gross margin +140bp, and LTM FCF/share from 0.1 to 17.2 since Q4 2022 5/ Neither $PLTR nor $SPOT funds the frontier. Both sit downstream of it, which makes them relatively immune to hyperscaler capex risk. $META shows the other side: $34.9B of capex in Q3, negative FCF, a $135–145B capex guide for the year, and now net debt. Over two years, LTM FCF/share went $META −44%, $SPOT +74%, $PLTR ~4x 6/ Yet $META’s core proves the point. AI trained on data only Meta has lifted time spent to records, app ad conversions +9.6%, price per ad +14% and ad revenue +26%, while cutting per-unit content-processing costs by more than 80%. FCF/share isn’t falling because of the core. It’s falling because Meta also chose to own the frontier 7/ My conclusion is that companies owning the data and running this playbook are set to compound FCF/share remarkably. $PLTR and $SPOT prove it, and $HIMS and $DUOL, my next two big bets, are built the same way. Whatever happens at the frontier, $PLTR wins: it’s the diffusion layer if closed models keep a premium and the model factory if intelligence commoditizes. I believe Palantir is set to become the world’s converter of tokens into value
15
15
3
194
26,137
“I think we’re just 1% into this AI wave and how we can leverage it at $DUOL.” Severin Hacker Google, Amazon, Netflix, Spotify and Meta built world-class business machines benefiting from a fundamental technological transformation unfolding at a particular moment in history. $DUOL is following the same path with AI time.com/collection/executiv…
1
1
30
2,110
🟢 BehavioralEdge 🟢 retweeted
Once AI can read the proteome, the human condition changes forever. $NAUT CEO Sujal Patel and CFO Anna Mowry join me on the podcast to discuss what it takes to get there. We record November 2. Community members watch first.
35
55
9
737
43,795