James Murphy

“Noise destroys information by introducing uncertainty” - Principles of Neural Design @whatishealth21 We have a problematic amount of noise online around health, with zero accountability by those who create it. It’s harmful to a healthy system, society. More to come.

Marc, this is like saying getting bloodwork causes heart disease. If the system is running great, you probably don’t go in for diagnostics. But if you’re looking to upgrade the system, a diagnostic and feedback process is a pretty good approach. The problem is in the execution (like most our systems in modern healthcare). Flawed foundational understanding. Poor mechanism of action. Low quality of care in interventions. Medicine has a shit track record with health disease right now. But you wouldn’t say stop getting blood work - And no one thinks bloodwork is the solution. You’re conflating terms. Introspection, done well, is a diagnostic and feedback tool. Not a solution or mechanism of action on its own.

My big conclusion from this week: Introspection causes emotional disorders.

Stop what you are doing and watch this

Reading the actual paper, the benefits seem to peak more at 1.5-2 cups per day and 150mg caffeine. They use 8oz serving size for coffee and don’t distinguish caffeine content between coffee servings and tea (tho tea has half as much). 2-3 cups is a misnomer for what people actually consume in this country. They also glaze over the fact that it seems the tea drinkers performed better than the coffee drinkers on reduced incidence of dementia by a wider margin than caffeine vs not drinkers.

More is not necessarily better. The benefits peak at about 300mg of caffeine, or 2-3 cups. After that, your liver enzymes (CYP1A2) saturate. Drinking more won't give you more protection and it just leads to jitters and poor sleep, which is bad for your brain.

A 43-year Harvard study of 130,000 people just dropped some major news about your morning coffee. It is more than a stimulant. It's the best way to prevent dementia. There is also one big catch though. Here is what the JAMA study found: (1/11)

@aakashgupta Can attest to this!

The neuroscience here is more radical than people realize. What you’re watching is a hormonal phase transition. Within minutes of skin-to-skin contact with a newborn, a father’s endocrine system starts a cascade that rewires his brain for the next 20 years. Testosterone drops 34% on average. Gettler’s 2011 landmark study at Notre Dame tracked 624 men and found that the ones who spent 3+ hours per day in direct childcare had the steepest declines. This matters because testosterone and parental sensitivity are inversely correlated. Lower T predicts more responsiveness to infant cues, more physical touch, more synchrony with the child’s emotional states. Meanwhile, oxytocin surges 33% above non-father baselines. Prolactin spikes. Estradiol rises. These are the same hormones that activate in mothers during pregnancy and breastfeeding. The entire “maternal bonding” cocktail fires in fathers through a different delivery mechanism: proximity and touch. Here’s where it gets wild. Dr. Pilyoung Kim at the University of Denver scanned fathers’ brains at 2-4 weeks postpartum and again at 12-16 weeks. The regions linked to attachment, empathy, and threat detection showed measurable increases in gray and white matter. The brain physically bulked up in areas responsible for protection and caregiving. And in mice studies, neurogenesis (new neuron formation) occurred in father brains within days of their pups being born. But only in fathers who stayed in the nest. The ones removed on day one showed zero new neuron growth. Physical contact was the switch. So the claim about brains being “literally rewired for protection” actually undersells it. The father’s brain grows new tissue. It shifts its entire hormonal architecture from mating optimization to caregiving optimization. The reward circuitry that previously activated for sexual stimuli redirects toward child faces and infant cries. The biological mechanism for fatherhood is one of the most aggressive neuroplastic events in the adult male lifespan. And it’s entirely dose-dependent: more contact, more holding, more time in proximity = stronger the neural and hormonal shift. That first hold is a pharmacological event.

@mariots 👏🏻👏🏻

In ~10 years, Oscar went from non-existent to neck-and-neck being the largest ACA health insurer in the country, growing revenues at 61% to $18.7-19B into 2026, and increasing market share in our footprint to 30%. Incumbent insurers are mostly useless. But better, consumer-driven healthcare still requires value-based care, risk management and other capabilities insurers are supposed to have. So we won't get better healthcare without a better individual health insurance market. And AI will transform healthcare - but it will be inflationary, not deflationary, if we don't turn US healthcare into a consumer market that rewards value, member experience and better outcomes. Oscar's results are showing yet again that even in the most turbulent of environments, we have figured out how to do that: we build better healthcare with better value for your money, and people are voting with their feet and choosing us. We won't stop until we've helped transform the rest of the healthcare system.

Seems like the authors might have missed an even bigger story in this. Can’t see all the data because it’s glossed over, but it looks like the reduced incidence of dementia in tea drinkers vs coffee drinkers was more pronounced than the caffeine vs not story. What am I missing?

Moderate consumption of caffeinated coffee or tea was linked to reduced #dementia risk and modest improvements in cognitive outcomes; no benefit was seen for decaffeinated coffee in an observational study of US adults. bit.ly/4amtd1c

This looks to be a pretty big study on caffeine and cognitive function. JAMA. 130k people, 48 years, observational study - so not causation. What’s getting me is how this is getting positioned vs what the data shows… The summary at the top says: “Higher caffeinated coffee intake was significantly associated with lower risk of dementia.” That sounds like the more caffeine, the better. But that’s not the actual story. Here’s from the conclusion: “Notably, the strongest associations were observed at moderate consumption levels; there were no additional advantages observed at higher intake levels.” The best cognitive outcomes were seen at ~2, 8oz cups of coffee or tea per day. Wait, 2 8oz cups? I don’t think there’s a coffee shop in America that serves an 8oz cup of coffee. And tea has half the caffeine as coffee. So this study is showing that probably sub 150mg caffeine / day is optimal. And further, which they seem to gloss over, the tea drinking cohorts were observed to have a 25-35% lower incidence rate of dementia compared to the coffee cohorts. Which in fact appears to be a BIGGER reduction than the 10-20% lower incidence rates of dementia this study focuses on comparing drinking no caffeine vs moderate/heavy caffeine.

@JamesonCamp Yes and yes. Also I see in the thread talk of scaling retail. If they can scale DTC first, focus there, the cash conversion cycle is much better.

I have a client in CPG who is bootstrapped and likely going to go from $5M last year to $15M this year (it’s doable, they did a 5x last year - can’t keep up with demand) They want to push for $20M, I’m telling them that unless the raise - it’s going to get super dangerous That cac is going to rise, supply chain will need to scale, money will need to come from somewhere etc Anyone have examples of brands who slowed down a bit as they grew at it was right decision?

@MaalkMan 👊🏻👊🏻

MALK will end 2025 +55% on the net revenue line and +130% on the EBITDA line vs last year. Plan for 2026 is +45% and +100% respectively. LFG

We’re a week out from Thanksgiving and I’ve already seen half a dozen “early Black Friday” sales by CPG brands. I don’t get it. I don’t get the lack of marketing creativity. I don’t get the idea of focusing team/advertising efforts on price instead of the value of the product. I don’t get taxing the brand teams during the holidays. I don’t get running sales that eat all (or most) contribution margin dollars. I don’t get eroding long term price integrity with your customer for a quick pop. I think about the amount of talent and effort each year that goes into running sales, to earn non-profitable revenues, from customers intending to spend time with their families, run by teams who could also be spending the time away from the desk. It all seems so misaligned with our biology - yet it’s taken as standard operating in CPG.

Consumer reports continues to reinforce why no one puts credibility in consumer reports. Watch out for this continued wave of “scientism” by social media folks too. Sounds science-y… But isn’t. That salad you had for lunch with leafy greens, nuts, tomatoes, maybe a base of rice, has more heavy metals than these reports on plant protein. And all the hullabaloo of judging against “the standard” is comparing against California’s Prop 65 law. Set as a drinking water standard for highly bioavailable lead from pipes at 1,000x below the then established “no observable impact” level. Yes, California found a level of “no impact” and arbitrarily set the standard 1,000 times lower. Pay no mind to the fact that eating vegetables trips you over that level. The EPA and FDA have never accepted those levels because, in their own words, “they are technically in-feasible” We clearly have a health and food quality problem in this country. But we also have a digital alarmism problem - which is making the first issue worse as we’re confusing people as to where to direct their attention.

@shaner_bud @AdamBreneman81 @drinkLMNT Stay salty! 💪🏻

@AdamBreneman81 Could care less about Franklin/PSU but love seeing the @drinkLMNT shout out💪🏻 @jamestmurphy_

My reaction to Penn State firing James Franklin

@moninvestor Have you taken a look at $INOD? 5 of mag 7 have contracts with them. $2.5bn market cap. Profitable. Data structuring company for AI models.

I’m not completely sold on the new stock yet, but I’m sharing this to show my thought process. I bought first, then started doing deeper research. I like the sector, the founder, and the fact that it’s still under a $1B market cap. When I looked into the leadership team and saw that it included people from across the tech industry, that was enough for me to take a position and then continue my research. My focus right now is RR & UAMY.

Great start to the week. - $IREN +14% ATH. $ONDS, $CIFR & $RR closed green. - $UAMY +4%. Need more time for write-up. - Bought a new stock. x10 potential👀 > Under $1B mcap, founder led & AI sector. > Team from AMD, INTL etc. > Researching. Could become LT hold. Nice day. 🤖

Help me understand their business. They say they have lots of accounts and distro… but earn $1.5MM per quarter?? Trade at $1bn valuation on a $6MM revenue run rate. That’s 160x revenues. Looks like a services business with robotics in the title. Genuinely asking, what am I missing?

$RR It’s great to see large accounts finally taking notice of this company. When I bought it, hardly anyone knew about it. That’s exactly why I love buying companies before they become famous. Now that bigger accounts are paying attention, it makes me confident the stock can reach my price target of 10x my initial purchase within the next 12-18 months.

@jordantwestecom 🙏🏻👊🏻

@jamestmurphy_ @jamestmurphy_ I love your product btw. Drink it every morning before my morning hike! Thanks for making it! 🙏

Slim & Salty. First available to our wholesale partners and key retailers.

@RossMackay111 Looking sharp! 💪🏻

The Cadence NYC Fuel Shop This city has some of the hardest working athletes in the world, and that’s who we built this for. I’ve poured everything into Cadence to make products that truly fuel you. Now, for the first time, you can walk into our own space in NYC, try the Core Fuel Gel, and grab the exclusive NYC Marathon Race Bottle alongside all our products. 📍 2 Rivington St 🗓️ Open every day until Sept 28 ⏰ Mon–Sat 10–6 | Sun 11–5 We work for you - the athletes of this city. Excited to see you in here.

Meta’s acquisition of scale AI gives a decent indication of value here IMO. time.com/7294699/meta-s… The non-meta customers then seeking alternative platforms. Turing and Handshake (plus INOD) for example. I don’t think the risk is concentration or outsourcing. I think both are tailwinds. Question in my view is whether this “older” company can adapt at the same rate as these new companies that have pivoted into the space.

@Andrecabv No growth for 30 years is a tough sell man. But yes maybe the stars are aligned. Honestly my bigger concern is customer concentration. And is this something a Mag7 can learn to do vs. outsource? I don't know enough about the complexities here.

As promised, I've spent some time researching Innodata as a potential new position. I'm not a wall street analyst, so this is a bit of my thought process as I review a new ticker. Here's a brief overview of what I've found. Why did $INOD interest me? - The stock is up 52% in a month and currently sits at #9 on IBD50 list. - In May, it was recognized by Wedbush as being one of 30 companies defining the AI revolution. - It's a $2B market cap (right about where I bought $RKLB and $HIMS). I LOVE investing in companies in the $2B-$5B range. What does $INOD do? Innodata is a 35-year-old data engineering firm focused on the red-hot field of AI data services, helping companies build and deploy artificial intelligence solutions. They handle cleaning, structuring, and labeling the messy raw data (i.e. text, images, audio, video) that AI companies and enterprises need to train models and run digital operations. Essentially, they help large enterprises deploy large-scale AI models. They are NOT an AI software company. They are a "picks and shovels" AI infrastructure play - Data Engineering/Operations and Consulting Services. Who are their customers? $INOD touts that they have 5 Mag 7 clients however in 2024 "one customer" accounted for 48% of their revenue. This presents a good amount of risk in my opinion. And we're not talking $100Ms of revenue streams from these megas (yet anyways). What about leadership? Jack Abuhoff joined the company in 1997 as President and CEO. Not technically a founder but has been at the helm for 30 years. I candidly need to learn more about Jack. I've watched a few of his interviews and he sounds incredibly sharp. But he does sport a weird man bun/pony tail from what I can tell - so another red flag for me (sort of kidding). Joking aside, I do not love that Jack's been in charge of $INOD for 30 years and hasn't been able to grow the business beyond a microcap status. Who are their primary competitors? Appen seems to be the most similar company in this space. Veritone another. Both struggling to grow while $INOD is growing quickly. What about finances? $INOD is profitable already with no debt. $60M cash on hand. Healthy. 2025 revenue growth is projected to be +45% (raised last ER) and last quarter was $58.4M (up 79% YoY). Gross margin 41% and profit margin at 18% look good. PE at around 50 and PS at 8.5 which feel appropriate for a profitable growth stock. $INOD cites a $200B TAM by 2029. So what do I think about $INOD as an investment? Honestly I feel mixed here. I like at the company is growing fast, profitable and financially stable. Also like that at a $2B there's a lot of room to grow. I've made a lot of money in AI picks and shovels ($NVDA and $NBIS) so their niche AI play is interesting. On the flip side, it's hard for me to get excited about data cleansing as a long term growth stock. I like investing in stocks that have some inspirational component to what they do. I also don't love that the CEO has been in this role for 30 years and they're at an all time high at a $2B cap. Insider ownership is only 5% so not a lot of internal alignment to move the stock price. I have a small position for now and the chart looks great. So I'm likely going to hold for now. I'm curious to learn more but will likely look for better opportunities. It could absolutely go much much higher. But I'm not all in on this one. Hope this helps. NFA DYOD.

@immad Email filtering that learns from among my own team to filter & prioritize. Ie, 5 teammates got spammed, one teammate sends to junk, the rest get flagged. Keeps data/IP local to company. Prob not venture scale - but noise is so high in email RN!

Respond to this tweet with a startup idea. I will tell you whether I would take a investor meeting or not and why for that idea, assuming it came in from a strong warm intro. I can tell in 280 chars whether an idea is interesting to me and why. Maybe it will be useful feedback.