Codes of Culture | Issue 98
The billion-dollar moment for women's health
Welcome back to Codes of Culture. I’m Ashumi Sanghvi.
Yesterday I attended a talk by Intelligence Squared with Demis Hassabis, the founder of Google DeepMind and Sebastian Mallaby, the author of The Infinite Machine, about Demis’s life and work. I shared more on LinkedIn here.
In today’s issue, we cover Midi Health’s billion-dollar valuation and what it signals for women’s health, Anthropic’s apparent move into app creation inside Claude, Waymo’s planned London launch, Google’s push to turn Gemini prompts into reusable browser behaviour, and the rise of functional fragrance as scent shifts from indulgence toward efficacy.
If you are new here, or want to catch up on the best of Codes of Culture, we’d recommend you start here and remember to subscribe for full access to our news, insights, podcast and global events.
📖In this issue:
Midi Health crossed the $1 billion mark, turning menopause care into a platform play across women’s health.
Leaked screenshots suggest Anthropic is building an app creator inside Claude, pushing the model layer directly into product creation.
Waymo is targeting a commercial launch in London by September, bringing autonomous ride-hailing to Europe.
Google turned Gemini prompts into reusable browser skills, embedding AI workflows directly into Chrome.
Functional fragrance is moving toward supplement logic, with scent increasingly framed around mood, stress, and focus.
1. CONSUMER HEALTH AND CAPITAL
What’s happening: Midi Health, the US women’s telehealth platform founded by Joanna Strober and Sharon Meers, raised $100 million in a Series D led by Goodwater Capital, pushing its valuation past $1 billion. The business began with menopause and perimenopause care and later expanded into cardiology, metabolic health, obesity, sleep, and longevity. It now serves more than 25,000 patients a week across all 50 US states.
TLDR:
$100 million Series D at a $1 billion-plus valuation. Goodwater led the round, with Serena Ventures, GV, Foresite, McKesson, and Felicis participating. Total funding is now $250 million.
Midi used menopause as the entry point, then expanded into broader care categories. The wedge was specific. The business was always larger.
AI is being used across chart analysis, clinical decision support, and protocol refinement. That operational layer is central to scale.
Women’s health continues to receive a small share of private healthcare investment. Midi joining Maven at unicorn status marks a shift in how the category is being capitalised.
The executive bench now looks built for scale. The company is being shaped for long-term expansion.
Why it matters: Women’s health is being financed with a level of seriousness that reflects the size of the market rather than the legacy assumptions around it. The more important signal is structural. Midi did not stop at a single life stage or one care need. It used menopause as an entry point into a broader care platform, then built the operational layer required to deliver that at a national scale.
2. AI INFRASTRUCTURE AND PLATFORM STRATEGY
What’s happening: Leaked screenshots on X appear to show Anthropic testing a full-stack application builder inside Claude, with tools to generate apps directly from a prompt. The interface appears to include live preview, database support, authentication, and one-click deployment, all within the chat environment. Anthropic has not confirmed the feature, but if authentic, it would mark a move beyond model access and developer tooling into direct product creation.
TLDR:
The leaked interface suggests Claude could generate front-end, back-end, database, auth, and deployment components in a single workflow. That points to product creation, not coding assistance.
This is distinct from Claude Code. Claude Code is a developer tool. The leaked builder appears aimed at users with little or no technical fluency.
Lovable built a major business on top of Claude. If Anthropic moves directly into that territory, the platform boundary changes.
Anthropic has already moved into higher-value product categories. This would extend that logic further up the stack.
The real question is not whether model companies expand upward. It is how much of the application layer they choose to own.
Why it matters: The stack is starting to compress in plain sight. Once the model layer can generate, deploy, and manage applications inside the same interface, value begins to concentrate higher up the chain. That changes the terms for every startup built on top of foundation models. Defensibility is increasingly a function of workflow, distribution, brand, and proprietary data rather than access to the model.
3. AUTONOMOUS MOBILITY AND INFRASTRUCTURE
What’s happening: Waymo, Alphabet’s autonomous vehicle company, is targeting an April passenger pilot and a September launch of commercial driverless ride-hailing in London, subject to regulatory approval. Around two dozen vehicles are already mapping the city in a supervised phase. If approved, London would become Waymo’s first commercial market outside the United States and its first in Europe.
TLDR:
April for a pilot, September for a commercial target. The key variable now is regulatory approval.
London offers a dense urban test case with complex traffic, heavy pedestrian flow, and demanding road conditions.
The UK’s liability framework matters. Under the Automated Vehicles Act, responsibility sits with the operator rather than the passenger.
Wayve is running parallel trials in the same city as Uber. London is becoming a live proving ground for autonomous systems.
The UK is treating autonomous mobility as part of its industrial strategy and transport policy.
Why it matters: London provides autonomous ride-hailing with a serious regulatory and urban testbed in Europe. The value in that is not confined to the vehicle or the passenger experience. It sits in fleet operations, insurance design, charging infrastructure, public policy, and city-level partnerships. Once autonomous mobility enters a market like London, the story becomes one of systems coordination as much as technology deployment.
4. CONSUMER AI AND DISTRIBUTION
What’s happening: Google launched Skills in Chrome, allowing users to save Gemini prompts as reusable workflows that run across webpages and open tabs. A prompt for comparing products, extracting recipe data, or summarising documents can now be saved once and triggered again through a slash command. Google is also releasing a library of prebuilt Skills across research, shopping, budgeting, and productivity.
TLDR:
A saved prompt becomes a repeatable browser action. One of consumer AI’s main frictions starts to disappear.
The early use cases are practical: product comparison, long-document scanning, recipe extraction, and budgeting.
Chrome’s scale makes this meaningful. A browser-level feature can quickly alter user behaviour.
As prompts accumulate inside a Google account, the browser starts to function as a personal instruction layer.
The browser is becoming an AI surface, not just a place where AI tools are accessed.
Why it matters: AI is moving from a destination product to an operating layer. Once instructions live inside the browser itself, discovery begins to shift before a user reaches a brand interface, a publisher page, or a product listing. That changes how information is filtered, how products are evaluated, and where intent is formed.
5. LUXURY WELLNESS AND FRONTIER BEAUTY
What’s happening: Business of Fashion reported that searches for “functional fragrance” reached a record high in March 2026, up 25 per cent from November and double the level of a year earlier. Brands including Initio Parfums Privés, The Nue Co., and Vyrao are defining the category by framing fragrance around mood, stress, focus, and physiological effect rather than sensory pleasure alone. What began as a wellness-adjacent niche is now moving into the wider fragrance market.
TLDR:
Search interest is rising quickly. Consumer curiosity is outpacing category definition.
The premise is straightforward: scent reaches the brain quickly, making it attractive for products framed around mood, stress, and focus.
The category is not uniform. The Nue Co., Initio, and Vyrao each position the logic differently.
Claims about stress reduction, mood, and focus are spreading across the fragrance category. The scrutiny on formulation and efficacy will increase with them.
Fragrance is being repositioned as both a functional and an expressive product.
Why it matters: Fragrance has traditionally been sold through identity, atmosphere, and desire. Functional fragrance introduces a different purchase frame built around outcome, effect, and intervention. That affects brand language, retail context, product development, and consumer expectation. Once a scent is sold with an efficacy promise attached, the rules around credibility begin to change.