55,000 therapeutic medical apps, almost zero independent evaluation. IEEE launched a voluntary verification registry: clinical benefit, technical soundness, data privacy compliance reviewed in six to eight weeks. Voluntary schemes have a self-selection problem worth watching.

Someone moved their entire Notion workspace into Obsidian using OpenAI's Codex. No importer, no lost tags. Just plain-English instructions on where everything should go. Coding agents earn their keep on the boring admin, not the creative work. https://dub.sh/XNjkahx

Dating apps optimise for time-on-app; Ateam is explicitly doing the opposite: no likes, limited scrolling, anti-ghosting design. Niche wellness dating with biomarker-tracking users as the target. Interesting design bet. Whether the addressable market is large enough is the open question.

Saw a demo where someone built a working app just by talking to a system, no code. I'm coaching someone on a medical ketogenic diet; no generic app fits. But build-by-conversation means the audience of one can finally build the exact tool they need.

Our relationships are one of the greatest determinants of health, happiness and longevity. We've optimised sleep, training and biomarkers; almost everything except how we connect. Ateam (NYC) rebuilds dating around workouts. The next longevity frontier is each other. https://dub.sh/tYuiHNb

When you exercise may matter as much as if you do. UK Biobank (92k adults): midday-to-afternoon activity beat morning routines for lower mortality. For adults with obesity, evening won (61% lower). Match timing to your chronotype. https://dub.sh/8qZ07Lg

Roughly half of the body’s VO2 max is determined by genetics. Age is also a factor. Improvements are nonetheless possible

Medicare reimbursing GLP-1s for cardiovascular prevention in patients without a prior cardiac event is the first genuine preventive reimbursement template for a new drug class. CAR-T at $400K+ per patient is nowhere near following that logic. The gap is economics, not biology.

Regenerative medicine's conference circuit makes sense once you follow the money: NewLimit's $130M Series B, XPRIZE's $10M x 10 Healthspan finalists, a longevity biotech market heading from $27B to $47B by 2033.

DeepSeek R1 briefly moved US AI stock prices in January 2025. That reaction confirmed open source model releases are now strategic instruments, not just community contributions. The asymmetry between Chinese and US lab behaviour on openness is the thing worth watching.

A 1950s Crest campaign sticks with me: it stopped selling cavity prevention (decades away) and sold fresher breath instead (tomorrow morning). Brushing became a habit. Proactive health has the same bug: we sell a healthy retirement to 25-year-olds.

Trail running's grace period is over. 2025 was the year the sport got too big to hide behind its dirtbag mythology. Now: who pays athletes, who runs the calendar, who polices doping. If GTWS, UTMB, PTRA and World Athletics can't align in 2026, the mirror keeps winning. https://dub.sh/GFHTA2g

CIOs on why they pick open source: productivity (86%), reduced vendor lock-in (84%), lower TCO (84%), faster innovation (82%), quality (79%), security (78%). It's no longer t-shirts and sandals railing against the corporate machine. It is the machine. https://dub.sh/FCk6Yir

Bart De Witte's 3-yr TCO for a 10-clinician practice: Local AI (Apple Silicon): €9,315 GPT-4.1 mini: €21,771 Claude 4 Sonnet: €75,673 Manus: €183,298 Frontier APIs overshoot clinical needs. SLMs run offline, keep patient data on-device, and can pass MDR cert. https://dub.sh/ZS4JM8K

St. Louis Fed: 43% of US workers use AI at work vs 32% in Europe. US workers also use it more intensively (5.2% of all hours, 2-3x Europe). Estimated cost: 3.2pp of productivity growth Europe didn't get. Same script as the 1995-2025 ICT gap. https://dub.sh/h6UafJ6

Maybourne has embedded a longevity nutrition menu into its main restaurants across six hotels; not spa-only, the actual dining room. Anti-inflammatory whole foods, developed with a functional medicine practitioner. The question is whether ambient food defaults move behaviour or just signal.

The longevity conference circuit is now 340+ events across 42 countries. The signal isn't the count, it's the clustering: October and June are jammed, July-August nearly dead. The hard question shifted from "is there an event" to "which one".

Clinicians worried about AI + patient privacy: run the models locally. Ollama or LM Studio, pair with a Whisper-based dictation app, start on non-clinical tasks. No cloud, no per-token bill, no third-party logging. https://dub.sh/n6J63Zd

Forever Healthy open-sourced AI4L 1.0: AI-generated reviews of longevity interventions, audited by independent agents against a 390+ item checklist with live citation verification. Loops until 100% pass. 'Audit-Driven Prompting' is the part worth stealing. https://dub.sh/iKeEpXE

Dubai averaged 83.6% in The Fit Guide's global boutique fitness rankings, scored on hotel-style criteria: reception, design, environment, class experience. The hospitality framing is the interesting part; most studios still treat the workout as the product, not the full experience.

Cross-industry partnerships keep beating same-sector ones. Stanley (century-old flask maker) x LoveShackFancy (floral fashion) sold out in hours. The distance between two industries is what makes the result surprising, and surprising gets noticed.

Rainy Sunday = LLM script execution comparison time

Enterprise AI transformation stalls not because of bad leadership but because the governance infrastructure doing its job correctly is structurally incompatible with fast experimentation. That's a design conflict, not a culture problem.

Vibe coding vs agentic coding is really a question about what happens when a prototype survives contact with users. One workflow has governance built in; the other doesn't. The risk isn't the tool; it's the drift.

Smart rings collect HRV, sleep stages, SpO2, and temperature passively and continuously. No screen, no notifications. Oura has been doing this for years; Samsung just entered with the Galaxy Ring; Apple may be next. The wrist is starting to feel like a constraint. #wearables #healthtech

GLP-1s: approved for diabetes, then weight loss. Now in trials for Alzheimer's, addiction, kidney disease, heart failure, and NASH. One mechanism. The breadth of targets wasn't in anyone's 5-year model. Every adjacent sector is repricing. #GLP1 #drugdiscovery

Abu Dhabi isn't pitching a longevity research park; it's pitching a live OS for healthcare, genetics, wearables and AI integrated into real clinical infrastructure. The appeal to founders is obvious. What's worth scrutinising is what governance model makes that integration speed possible.

WHOOP embedding clinician video visits and EHR integration via HealthEx is the wearable-to-clinical-workflow bridge people predicted for years. The friction: regulatory classification hasn't caught up. Fitness device or medical one? That answer changes everything about liability and data governance.

AI tools default to praise by design. 'Be brutally honest' is four words that cut through the sycophancy loop. The people not typing them get confident feedback on mediocre thinking; AI makes you worse without you noticing. #AI #criticalthinking

340+ longevity events tracked in 2026. European vs US conferences: different speakers, formats, topics, and sponsorship patterns. The data shows which side is moving faster and where the money is concentrating. #longevityevents #healthspan

Two decades of biological age research and the dominant use case is still self-tracking. Structuring it as a competitive cohort with defined winners, across brain, body, face, and clinical markers, is a different hypothesis about what drives behaviour change.

Most B2B partnerships don't fail from picking the wrong partner. They fail from what happens inside before signing anything. The misalignment was already there; the partnership just made it visible. #openinnovation #B2Bpartnerships

Prevention budgets down 28% in real terms; prevention strategies up in passion. Same period. That's not contradiction, it's structural lock-in: acute crisis always wins the resource fight in a demand-led system. Whether private consumer health models can break that is genuinely unclear to me.

Luffu aggregates health data across a household and learns family behavioural patterns. The founders built Fitbit. The interesting design tension: family health monitoring and individual data consent are not the same problem, and most AI health platforms haven't seriously addressed that gap yet.

The J-curve for AI automation is shallow and short. For augmentation it's deeper and longer. Most finance teams will choose shallow every time. Whether that's rational capital allocation or slow organisational decay dressed up as efficiency is the question I can't quite resolve.

AI has mastered the 95% of knowledge work ; the repeatable, structured, predictable tasks. What's left is the 5%: judgment, context, knowing what question to ask. That's where value is concentrating. Are you a 5 percenter?

Luxury is converging on longevity and biohacking as its new status layer. LVMH spending $300M+ on sports, Kith at $40K/year for a club, Aman building detox programmes; the product is now biological optimisation. Worth watching how that shapes who gets to participate in the wellness conversation.

Biological and chronological age are decoupling faster than most institutions have noticed. The 60-year-old with a 30-year-old cardiovascular profile is no longer the outlier. Real implications for how we design products, workplaces, and expectations.

Harvard study in Science: AI matched or outperformed physicians across 76 real ER cases on triage, diagnosis, and management reasoning. Not a simulation. Authors call for controlled clinical trials next. Reasonable ask; but the benchmark has already shifted whether trials happen or not.

The functional beverage market has a signal-to-noise problem. These five ingredients actually have science behind them: probiotics, adaptogens, collagen, yerba mate, whey protein. Each targets a different mechanism.

Traditional hackathons are constrained by human stamina. AI-agent hackathons aren't. Thousands of specialised agents, running in parallel, generating and evaluating millions of ideas. The bottleneck shifts from idea generation to quality control. That's actually the more interesting design problem.

Innovation valuation fails at the translation layer. The xV metric converts confidence, predicted value, time sensitivity, and strategic fit into one monetary figure on a rolling cadence. Treating portfolio value as a living number; not an annual snapshot; is worth stress-testing.

$3.74B into longevity biotech in Q1 2026. 173 AI-discovered drugs now in human trials. First FDA-approved mitochondrial therapy. The bottleneck has shifted from discovery to delivery ; the gap is now in clinical infrastructure, not the science.

Interesting inversion in AI-first org design: stop designing AI for your company, redesign the company for AI agents. The blocker isn't model capability; it's that most org structures were built to coordinate humans, not autonomous agent networks. That gap is where enterprise AI value gets lost.

The partnerships that look absurd on paper are often the ones that redefine categories. Went through five cross-industry collaborations that proved everyone wrong — and found a common thread in what made them work. https://dub.sh/0EKI9Vc #openinnovation #B2B

60% of UK workers value private health insurance over an extra holiday week. C-suite already buys its own longevity stack. Can HR rebuild benefits around healthspan before 42-year-olds notice? www.linkedin.com/pulse/you-ca...

The clinician isn't being replaced. The clinician is being re-intermediated. Five patients walked their own path to a diagnosis, a second opinion, or a treatment the system missed. hyperadvancer.com/blog/the-pat...

🎰 The AI just matched Dole with WHOOP in a universe-altering partnership that will single-handedly reshape the very fabric of commerce. Try one spin. The algorithm has spoken. roulette.hyperadvancer.com

New batch just dropped in Venture Roulette. The category: AI, longevity, fitness, healthcare. The vibe: a few of these will look completely obvious in 5 years and absolutely unhinged right now. Go find out which: roulette.hyperadvancer.com

A cardiologist. No dev team. Seven days between shifts. Third place out of 13,000 at Anthropic's hackathon. The bottleneck in healthcare AI has flipped. The people closest to the problem can now build the solution. hyperadvancer.com/blog/the-inv...