Apple explains the provenance of the models available through Foundation Models Framework. You don’t have to guess. The smaller on-device model is Apple’s own proprietary LLM, and the larger is distilled from Gemini. The larger PPC-hosted FMF models are fine-tuned from Gemini, as Apple stated when it announced the partnership. mobiledevmemo.com/unpacking-…

Ad-supported media generates enormous consumer surplus, providing billions of people with costless access to information, entertainment, and communication infrastructure that would otherwise be the exclusive remit of the wealthiest segment of society.

WWDC 2026: Implications for the App Economy To my mind, the new tools, features, or pieces of functionality introduced at WWDC 2026 that are meaningful to the app economy are presented in seven developer videos. Across these seven, the developer-facing announcements can be bucketed into two workstreams. The first is Siri and Apple Intelligence integration via App Intents: Apple is clearly nudging developers toward schema-based entities and actions, semantic indexing in Spotlight, on-screen entity annotations, and UI interaction declarations so Siri can understand app content, context, and follow-up requests with much less custom NLP work. The second is distribution and monetization: richer App Store creative placements, cancellation-time retention messaging, more expansive subscription packaging, group and organization purchases, and new game-delivery tooling around StoreKit and Background Assets through a deeper integration with Unity. mobiledevmemo.com/wwdc-2026-…

Unpacking Apple’s update to Foundational Models Framework Apple’s Foundation Models Framework now features a 20-billion-parameter on-device model using dynamic pruning to manage memory constraints. The update introduces CoreAI for third-party model integration and emphasizes Private Cloud Compute for privacy. Small developers receive free access to Apple's Private Cloud Compute-hosted models, framing Apple's infrastructure as a strategic alternative to external APIs. mobiledevmemo.com/unpacking-…

Bending Spoons has filed for an IPO, targeting a $20-22 billion valuation. The company’s roll-up strategy focuses on acquiring post-peak yet stable digital businesses and integrating them into its proprietary operating platform. Despite carrying significant debt, the firm demonstrates strong revenue growth, high subscription retention, and minimal reliance on paid user acquisition. mobiledevmemo.com/bending-sp…

I am hiring a part-time producer for the MDM Podcast. The primary responsibility of this role is to identify relevant guests and to build a robust, consistent pipeline of conversations about the digital economy. Each week, you'd surface people worth speaking to based on their recent work, domain expertise, research, writing, company-building, or other contributions to the space. The role also includes guest outreach, scheduling, and preparation. That means managing communications with guests, coordinating calendars, and preparing briefing documents in advance of recordings. This is a part-time, compensated position. In success, I imagine it would require roughly 5-8 hours per week, although I'm focused on outcomes and won't count or track hours worked. The goal is simply to maintain a consistent pipeline of high-quality guests and conversations. No prior podcasting experience is required. There is no editing, social media, or production work involved; I've automated that. This is fundamentally a research role. The ideal candidate is intellectually curious, closely follows developments in technology and the digital economy, and can identify the people best positioned to explain important ideas. To apply, use the Contact form on MDM to send: - A link to your LinkedIn profile. - A description of the prototypical MDM Podcast guest in <= 2 sentences.

This FT chart visualizes the dramatic increase in apps published to the iOS App Store. I wrote about this a few weeks ago, when SensorTower published similar data. I don't find the increase particularly interesting, since that's exactly what one would expect with the proliferation of AI-enabled coding tools. What I do find noteworthy is: - this is unique to the App Store; Google Play has seen no commensurate increase, since it tightened its automated submission review standards, also through AI: more output meets greater scrutiny, both empowered by AI. - the surge in apps published to the App Store coincides with Apple launching a new ad placement in App Store Search. As I note in the first episode of The Prosperous Society, AI-enabled output shifts the binding constraint for commercial success to distribution.

The Amazon DSP gained more share in Q1 2026 than any other enterprise DSP, and continues to be the catalyst for @AmazonAds growth. I had the pleasure of sharing why they've been so successful on the @MobileDevMemo podcast with @eric_seufert this week. Give it a listen here 👇👇 open.spotify.com/episode/32M…

Your phone isn’t secretly listening to you for ad targeting, part 2 Last month, the FTC fined Cox Media Group (CMG) and two other companies to settle allegations that the firms deceived customers about the utility of a marketing service sold by CMG, which claimed it could target ads based on conversations recorded by the microphones on consumers’ phones. The FTC’s complaint is unequivocal: CMG’s claims about the efficacy of its Active Listening product were utterly fabricated. mobiledevmemo.com/your-phone…

Meta’s subscriptions and the advertising narrative blind spot Public market investors understand subscription economics intuitively. They may be less fluent in AI-enabled advertising economics, which are harder to model, especially across the three axes I highlight in the Prosperous Society series. A recommendation-system improvement that lifts advertiser ROAS or user time spent by 3%, expands advertiser participation, and compounds into future auction density is much harder to conceptualize than a recurring monthly stream of consumer subscription revenue, despite potentially being worth orders of magnitude more. Meta already has what other chatbot operators don’t: a high-performance advertising platform. Adopting subscriptions feels like a clumsy attempt to adapt to investor expectations rather than reshape them. $META $SNAP mobiledevmemo.com/metas-subs…

Google revealed 1BN MAU for AI Mode at Google I/O, and Search grew 19% in Q1.

Must listen interview for anyone thinking about consumer AI & Ads, and of course the upside for democratizing access to AI. Lots of nuggets in there by @eric_seufert and @benthompson, great articulation of the principles and the direction.

As always, it was fantastic to join Ben Thompson on the @stratechery podcast. We discussed the generative revolution in digital advertising, Meta’s communications challenges, the evolution of Google Search, Apple, AppLovin, and the components of a Prosperous Society.

The generative RecSys revolution The legacy two-stage approach to RecSys comprises retrieval and ranking, and is largely implemented by scaled platforms in two mechanisms: the “two towers” framework, proposed by Google in 2016, and the Deep Learning Ranking Model (DLRM), proposed by Meta in 2019. But the modern generative retrieval framework reframes RecSys as a generation task rather than a scoring task: instead of retrieving and scoring a massive volume of potential candidate items for the best fit, the mechanism generates the next item’s semantic ID with a model. This concept might be instigating a fundamentally new paradigm in the RecSys and digital advertising ecosystem. mobiledevmemo.com/the-genera…

I'm happy to share some of the research that I've been working on today: As ad platforms become more opaque and automated end-to-end, advertisers are left with few levers of control over campaign performance. I wanted to interrogate an idea: could advertisers treat "black box" platforms as teacher models and use a behavioral distillation process, focusing on multimodal creative and ad context, to predict ROAS performance for ad instances? And if so, what kind of model could best express the interactions between those features? Most machine learning research focused on digital advertising is published by the largest platforms themselves. I wanted to address this asymmetry and explore what advertisers could build independently to improve performance within this increasingly automated environment. That idea ultimately became DeCANT: a Deep Creative Attention-based Network for pre-Testing. The model architecture uses self- and cross-attention to condition the semantic interpretation of a creative on the environment in which it is deployed. The model architecture supports multimodal ad creative and predicts ROAS on a context-conditioned basis. Operationally, a model like DeCANT can fit into a pre-testing regime as an automated filter: creatives are produced through a generative pipeline, and the student model that learns the process for the given advertising channel is invoked on the proposed creative and context. The model produces an expected ROAS, which is compared against the advertiser's testing threshold to determine whether the creative is uploaded to the platform. mobiledevmemo.com/introducin…

AppLovin publishes Gist, a social media app mobiledevmemo.com/applovin-p…

I wrote about this agency’s inflated claims, and why they allowed the misperception to persist that phones passively listen to conversations for the purposes of ad targeting, in September 2024, in a piece titled “Your phone isn’t secretly listening to you for ad targeting.” Now the company has been fined by the FTC for misrepresenting its capabilities. mobiledevmemo.com/your-phone…

AppLovin $APP has quietly published a Pinterest-like social app: Gist.

Google’s Gambit, Part 4: The end of Static Search The commercial opportunity as Search converges to a hybrid chatbot interface is fairly obvious: more queries (demonstrable, from Google’s disclosure), more ad exposures per Search session (given the conversational nature of chatbot engagement), and more direct integrations into checkout. And when the discovery and transaction interfaces converge into a single point of contact, retailers must not only compete for exposure but should, in theory, be willing to pay more for it, because the funnel compresses and the distance between discovery and the transaction shrinks. mobiledevmemo.com/googles-ga…