/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

A profile of Simile, which offers “agentic twins” modeled on real people to provide answers for polling and market research for companies like CVS and Gallup

AI startup Simile offers ‘agentic twins’ modeled on real people to provide answers for polling and market research

Wall Street Journal Belle Lin

Context & Ripple Effects

Simile’s profile follows its $100M emergence from stealth to predict human behavior, extending that positioning from purchase prediction into polling and market-research responses.

It also advances a use case explored in earlier coverage of digital twins built from real people for focus groups and other research. The involvement of CVS and Gallup puts the approach in settings where the credibility and provenance of responses matter.

First-order effects

  • CVS, Gallup and similar clients gain a synthetic-respondent option for generating answers in polling and market-research workflows, rather than relying solely on live participant panels.
  • Simile’s immediate challenge shifts from demonstrating that its models can converse to showing that responses modeled on people are useful for research decisions.

Second-order effects

  • Traditional research vendors and panel providers face pressure to distinguish live-sample data from modeled responses and to explain where each method is reliable.
  • Clients using such systems will need stronger controls around the source data, consent, and representation behind a twin, especially when results inform consumer or public-opinion conclusions.

Third-order effects

  • If adoption persists, market research could become a hybrid market in which human panels supply training and validation while AI agents supply faster scenario testing and iteration.
  • The model makes likeness governance a core industry constraint: durable adoption depends on credible rules for permission, compensation, transparency, and limits on using a person’s behavioral data.

The trend: AI is moving from generating content to simulating bounded human perspectives, bringing data provenance and likeness rights into operational decision systems.

Discussion

  • @renrut-mas Sam Turner on bluesky
    “We saw how poorly opinion polls and betting websites reflect public opinion and thought: we can make this so much worse.”  [embedded post]
  • @artisny @artisny on bluesky
    And just where are they getting the training data for these “agentic twins” modeled on real people, hmm?  Are we talking the internet generally?  Or: individual people whose privacy and personal data were violated—without their knowledge or consent or *payment*
  • @hypervisible.blacksky.app @hypervisible.blacksky.app on bluesky
    Agents “are essentially digital clones of real individuals, who are interviewed to gather their preferences, personality and other traits.  Simile then combines that data with participants' behavioral and purchase data to ensure 'generalizability and visibility into people's thou…