// attention economy

All signals tagged with this topic

AI startup Basata automates the doctor's callback, exposing healthcare's labor math

Basata is automating patient callbacks—a visible friction point in healthcare—by using AI for triage and scheduling. The model works until regulators or liability concerns force the question of who's responsible when an AI system misses something a human would catch. The startup's approach reveals that healthcare's callback problem isn't a staffing shortage but a profitability equation: clinics have optimized around minimal administrative labor, so a functioning callback system requires either hiring staff or deploying automation that shifts risk. This model depends on the healthcare system continuing to outsource accountability to startups rather than holding providers legally responsible for offloading clinical judgment to machines.

Half of Young Adults Get Health Advice From Influencers and Podcasters

The health information economy is now dominated by unregulated creators rather than credentialed sources, with Instagram and TikTok functioning as de facto medical authorities for under-50 Americans. Platforms optimized for engagement—not accuracy—determine what health claims reach millions, while pharma and supplement brands scale medical misinformation through affiliate relationships and creator sponsorships with minimal friction. Insurance companies are competing with wellness creators for patient behavior change, and the FDA's enforcement capacity cannot keep pace with the volume of claims distributed across short-form video.

Footwear Sales Flatline While Consumers Resist Price Increases

The shoe industry faces demand stagnation, not contraction. Consumers are resisting margin expansion. Nike and On have discovered their pricing power has eroded; flat unit sales at higher average selling prices mask an inability to grow the category. Footwear has become a mature, replacement-driven market where consumers comparison shop and resist premium positioning. The competitive advantage shifts from growth narratives and innovation to supply chain efficiency and inventory discipline, favoring operators with scale over aspirational challengers.

Looksmaxxing replaces thinness as the new body ideal

The shift from "thin" to "looksmaxxed" moves away from a single body type toward individualized aesthetic optimization—people sculpt whatever version of attractiveness suits them, whether that's muscle, curves, or symmetry. This fragments the diet-industrial complex: instead of everyone chasing the same silhouette, the market splinters across personalized fitness routines, cosmetic procedures, skincare stacks, and social media coaching, each targeting a specific "type." The monetization opportunity expands rather than contracts, since looksmaxxing demands continuous investment across multiple categories rather than just calorie restriction.

Tesla's 2017 Plan to Launch Rival AI Lab, Newly Revealed

Internal messages between Shivon Zilis and Tesla leadership reveal a 2017 strategy to build a competing AI operation anchored by Sam Altman or Demis Hassabis. Had the plan succeeded, it would have altered the trajectory of both Tesla and OpenAI. The disclosure reshapes the competitive history of the 2010s: rather than separate institutions pursuing distinct paths, internal power struggles and executive poaching attempts determined which organizations led AI development. It also shows that executive mobility and capital concentration—not just technical talent—decided AI leadership. Zilis and Tesla's pursuit of Hassabis or Altman suggests that access to specific individuals, not labs or methodologies, drove valuations and competitive advantage.

Smart TVs Are Quietly Building Surveillance Infrastructure in Your Home

Samsung and LG are already extracting visual data from living rooms at industrial scale—Samsung at 7,200 frames per hour, LG at 360,000—ostensibly for ad targeting and content recognition. The infrastructure they're building is surveillance-grade capture capability that vastly exceeds what current monetization requires. The gap between what these companies need to collect and what they are collecting points to either aggressive future use cases (biometric analysis, attention tracking, household composition profiling) or a technology-first approach where collection precedes permission and justification. For consumers, this means the living room is being enrolled in a data extraction pipeline without meaningful consent mechanisms or transparency about what "batching uploads every 15 seconds" actually contains.

Google's Uninvited 4GB AI Download Crosses the Line

Google has begun installing a 4GB AI model on Chrome users' machines without explicit consent, embedding computational weight into consumer devices to train its generative capabilities at scale. The installation arrives as a browser update, not as a feature users can opt into or decline. The move treats user devices as extensions of Google's compute network, prioritizing AI training speed over transparency. It gives consumers a concrete reason to switch to Chromium alternatives or competitors that haven't made the same choice.

Apple Finally Lets Users Build Their Own Wallet Passes

After a decade and a half of gatekeeping digital wallet functionality, Apple is surrendering control to end users. Its developer ecosystem failed to deliver the breadth of pass types consumers needed. This moves friction from "convince Apple to add support" to "figure out the format yourself"—democratizing wallet innovation but risking fragmentation across amateur-built passes of wildly different quality. The shift reflects consumer demand for customization over curation, particularly in categories where Apple's roadmap lagged (loyalty programs, local transit, regional payment schemes) and third-party apps couldn't legally compete.

Twitch Legitimizes "Mogging" as Streamers Weaponize Comparison Culture

Twitch's rule change permitting streamers to directly compare and mock each other's appearances or performance on-platform formalizes what was already happening in clips. The move is an explicit bet that conflict-driven content generates more engagement than community guidelines historically allowed. Creator economics have inverted moderation priorities: platforms now optimize for the viral moment over the safe space. Twitch is legalizing the dunking behavior that drives clips, which drives algorithm placement, which drives sponsorship valuations. The infrastructure rewards a creator hierarchy built on public humiliation.

Google Quietly Installs 4GB AI Model on Chrome Desktop

Google is embedding generative AI directly into Chrome's client-side infrastructure, shifting computation from cloud servers to individual machines. The move democratizes access to Gemini Nano while making the model harder to audit or control at scale. It mirrors how browsers became the dominant OS for consumer software, except the stakes now involve training data collection, model behavior, and surveillance surface area that lives on your hard drive. The "silent installation" framing obscures a significant change to device ownership: users inherit storage, computational load, and security responsibility for Google's infrastructure without explicit consent or clear removal pathways.

Google tries to salvage publisher traffic after AI Overviews decimated clicks

Google's AI Overviews are cutting publisher traffic by 58%—clicks that once went to websites now stop at Google's summary layer. The "Further Exploration" section nominally addresses this but doesn't restore lost traffic; it creates a secondary tier that publishers must now compete harder to appear in. This exposes a structural conflict in Google's model: AI summaries improve search engagement and ad placement but damage the publisher ecosystem that supplies Google's content. Publishers face a choice between optimizing for this new distribution layer or accepting traffic loss.

Google AI Search Now Pulls Real-Time Voices From Reddit and Social Media

Google is outsourcing credibility signals to user-generated platforms, betting that forum discussions and social media posts will outperform its algorithmic ranking of traditional publishers. This threatens the SEO playbook of the last 15 years—brands can no longer rely solely on optimized website content to win visibility, since Google now gives equal real estate to Reddit threads and TikTok posts. For certain query types (product recommendations, advice, lived experience), consumers trust peer networks more than institutional sources. Brands must either build community presence on these platforms or watch their search authority shift to crowdsourced alternatives.