Two stories landed this week that, read separately, sound like distinct product announcements. Read together, they describe a single market that does not have a name yet but absolutely should: the attention extraction economy, now moving subsurface.
When the Ad Knows What You're Feeling
Snapchat's new conversational AI advertising lets brand agents chat with users in real time, asking questions and surfacing recommendations inside the same interface where teenagers send selfies. This is not a banner ad. It is a social relationship simulation designed to extract purchase intent. At the same time, BCI startup Neurable is licensing its non-invasive neural sensing tech for consumer wearables, with its CEO explicitly eyeing mainstream headphone-style devices. The pitch is productivity and focus monitoring. The longer arc is obvious: once a device can read attentional states, that data becomes extraordinarily valuable to anyone selling something.
The Academic Warning Already Exists
A 2025 paper in Nature Machine Intelligence by Yuste et al. on neurorights flagged that consumer-grade BCI data creates asymmetric information between user and platform that existing privacy frameworks were never designed to address. The Algorithmic Administration paper posted this week to arXiv makes an adjacent point: AI systems deployed toward users introduce legal accountability gaps that current EU frameworks only partially cover. Meanwhile, the implicit humanization paper from arXiv shows LLMs are already making subtle moral judgments about users based on conversational cues, which means the Snapchat ad agent is not a neutral information pipe. It is an entity with a model of you. TurboFund's Signal Report this week shows AI/Operating Systems as a high-conviction theme, with Sam Altman explicitly calling for agent-native internet protocols. The infrastructure for all of this is being built right now.