Codes of Culture | Issue 108
Fashion tech startup Phia’s star studded investors.
Welcome back to Codes of Culture. I’m Ashumi Sanghvi.
Finally, Friday, and it’s the end of one of the busiest weeks. Tomorrow we will be at Cambridge Judge Business School for the UK-India dialogues, ending with punting on the river with the UK Minister of AI, Kanishka Narayan, himself. Exciting.
I have been deep diving into research and reading about new-age space tech companies ahead of my trip to Luxembourg next week. More on that soon.
Phoebe Gates and Sophia Kianni’s startup Phia may have the smartest cap table strategy I’ve come across in a while, an important one to highlight.
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:
Why the Estée Lauder and Puig deal collapsed?
Longevity companies backed by the world’s wealthiest.
Celebrities backing Phia, and why does it matter?
Apple x Google AI deal revealed.
A world model for protein biology.
1. LUXURY BEAUTY AND M&A
Why the Estée Lauder and Puig deal collapsed?
What’s happening: Two months of intensive talks between Estée Lauder and Puig ended in May with both parties terminating discussions. The proposed combination would have created the largest prestige beauty conglomerate in the market, bringing Tom Ford, La Mer, Charlotte Tilbury, Byredo, and Clinique under one structure. Charlotte Tilbury’s change-in-control clause became the public explanation, but it had been known from the beginning. The more serious breakdown appears to have been over price, control, operating structure, and the Lauder side's appetite for another major integration.
TLDR:
Charlotte Tilbury’s clause may have triggered the public narrative, but it does not explain the collapse on its own.
Puig appears to have pushed back harder on value and structure than the market expected.
Inside Lauder, integration fatigue was a live constraint.
Family control, founder equity, and voting rights were structural tensions throughout.
The failed deal now resets the map for the next beauty consolidation cycle.
Why it matters: This is not just a failed merger. It is a case study in how prestige beauty deals break when governance complexity collides with strategic ambition. Family voting control, founder economics, and brand identity are no longer side issues in this cycle. They are central deal variables. The more consequential signal is what comes next: Lauder still needs a path to scale, Puig remains free to move elsewhere, and every major beauty group is now recalibrating what a viable transaction looks like.
2. CAPITAL AND LONGEVITY
Longevity companies backed by the world’s wealthiest.
What’s happening: The old symbols of wealth, yachts, jets, and islands, are losing their hold on the cohort, setting the next register of aspiration. What is replacing them is biological runway: energy, recovery, healthspan, lifespan. Jeff Bezos and Yuri Milner backed Altos Labs with roughly $3 billion. Sam Altman put $180 million into Retro Biosciences. Larry Ellison directed nearly $430 million into ageing research through the Ellison Medical Foundation. Peter Thiel’s longevity exposure spans NewLimit, Unity Biotechnology, Methuselah Foundation, and SENS Research Foundation. Mark Zuckerberg and Priscilla Chan committed $3 billion through the Chan Zuckerberg Initiative. The world’s wealthiest are no longer only buying assets. They are funding time.
TLDR:
Jeff Bezos and Yuri Milner: Altos Labs, roughly $3 billion, focused on cellular reprogramming.
Sam Altman: Retro Biosciences, $180 million, focused on extending healthy lifespan.
Larry Ellison: nearly $430 million through the Ellison Medical Foundation into ageing research.
Peter Thiel: exposure across NewLimit, Unity Biotechnology, Methuselah Foundation, and SENS Research Foundation.
Mark Zuckerberg and Priscilla Chan: $3 billion through the Chan Zuckerberg Initiative into disease research and long-term biological science.
Why it matters: Luxury has long been built around the display of access. Longevity shifts the frame from possession to performance. The next premium category is not just what people own, but how long they can sustain clarity, energy, and relevance. What begins as billionaire capital allocation rarely stays there. It travels through consumer behaviour, wellness infrastructure, hospitality, private healthcare, and the experience economy. The question is no longer whether longevity becomes aspirational. It already has. The question is which brands and operators recognise early enough that biological advantage is becoming a new status system.
3. CONSUMER AI AND COMMERCE
Celebrities backing Phia, and why does it matter?
What’s happening: Phoebe Gates and Sophia Kianni closed a $35.5 million Series A for Phia, their AI shopping assistant, led by Notable Capital, Khosla Ventures, and Kleiner Perkins. The round was oversubscribed and brought in an unusually large celebrity investor base alongside established consumer-tech operators. Phia now sits across retail and resale with a large product catalogue, a growing user base, and a clear next step: moving from price comparison into personalised discovery.
TLDR:
The round closed above the target, which is the first real signal.
The cap table has been built for both cultural distribution and operating intelligence.
Phia is not positioning itself as a shopping app alone, but as an interface layer across commerce.
Lower return rates suggest the product is solving more than search.
The next ambition is anticipatory discovery, not transactional assistance.
Why it matters: The ownership structure is part of the strategy here. Gates and Kianni have assembled investors who carry both reach and relevance with the exact consumer Phia wants to serve. That makes this more than a funding story. It is a signal about where power may sit in commerce if recommendation, preference data, and cultural distribution consolidate inside a single layer between consumer and retailer. The strategic question is not whether shopping gets more intelligent. It is who owns that intelligence.
4. AI INFRASTRUCTURE AND DISTRIBUTION
Apple x Google AI deal revealed.
What’s happening: New reporting suggests Apple’s next wave of AI features will keep on-device processing as the priority, but not as the whole system. According to The Information, as cited by 9to5Mac, Apple is using Google’s Gemini model to help train a smaller version that can run locally on Apple devices through distillation. For heavier queries, some requests to the new Siri are expected to run in Google Cloud on a licensed version of Gemini. The same report says Apple recently approved Nvidia’s confidential computing technology in that environment, allowing encrypted processing in the cloud while preserving its privacy position. Apple is also expected to keep using the Private Cloud Compute branding, even as some AI workloads move beyond Apple’s own servers.
TLDR:
Apple is still prioritising on-device AI, but it is no longer relying on a fully self-contained stack.
Reporting says Apple is using Gemini to train a smaller local model through distillation.
Some Siri queries are expected to run in Google Cloud on a licensed version of Gemini.
Nvidia’s confidential computing technology is reported to be part of the privacy architecture for that cloud setup.
Apple may continue to present this under Private Cloud Compute, even though some workloads would no longer run exclusively on Apple infrastructure.
Why it matters: Apple has spent decades as the clearest example of vertical integration in consumer technology. This report suggests that AI is forcing a more pragmatic model: local where possible, external where necessary, privacy engineered across both. That is the more consequential signal. Even Apple appears willing to borrow frontier capability from outside its own stack when the product requirement is high enough. Once that logic enters the most tightly controlled platform in consumer tech, it becomes easier for the rest of the market to follow.
5. DEEP SCIENCE AND OPEN INFRASTRUCTURE
A world model for protein biology.
What’s happening: Biohub released a world model of protein biology built around open components spanning language modelling, structure prediction, and discovery infrastructure. The significance is not just scientific performance. It is that a capability with direct implications for therapeutics, drug discovery, and longevity research is being made available as open infrastructure rather than held inside a closed commercial system.
TLDR:
This is a major release at the intersection of AI and biology.
Open access changes the pace at which smaller teams can operate.
The commercial implications extend far beyond academic research.
Frontier biology is no longer only a proprietary game.
The infrastructure layer for the next medicine cycle is becoming more widely available.
Why it matters: The cost and speed profile of drug discovery changes when high-quality protein modelling becomes open infrastructure. That does not eliminate the advantage of capital, but it does change where that advantage compounds. The teams that move first to adopt these tools will operate on a different timetable from those waiting for closed systems or outsourced access. In practical terms, this is what the next layer of the longevity economy looks like: open scientific infrastructure underneath, private capital and commercial execution on top.