Posts & Deep Dives
Transparent explanations of how Affinity Atlas works, critical takes on the dating app industry, and honest analysis of the competitive landscape.
Show Your Working
The feed is an interface for engagement. The match card is an interface for trust. Explanation-first match cards are becoming normal - and "why" will become as important as "who".
Trust at Scale Requires Proof
"Verified" is drifting from a badge to a prerequisite. That will reduce some harms. It will also create new ones - exclusion, privacy risk, and a verification arms race.
Designing Against Addiction
If your unit of success is time-on-app, you will build swipe addiction. Here are concrete design patterns and metrics for connection-first dating products.
AI in Dating: Assistant vs Companion
Most "AI in dating" debate misses the core distinction: tools that point you toward real-world connection vs companion systems that manufacture "intimacy primitives" and can create dependence.
The Right to an Explanation (in Dating)
Dating algorithms decide who you see, who sees you, and how you are ranked. A score without a reason is power without accountability. What meaningful explanation should look like.
Dating Apps Are Sensitive-Data Systems First
Dating apps sit at the intersection of identity, location, sexuality, and intent. Treat them like normal social products and you will build the wrong privacy model.
The Dopamine Loop
How dating apps hijack your brain's reward system with variable-ratio reinforcement and slot machine mechanics. The neuroscience of why swiping is addictive even when it never leads to a date - and what non-addictive design looks like.
What Your Spotify Says About Your Compatibility
Music taste is one of the strongest personality proxies available. How Affinity Atlas uses Spotify listening data, Big Five personality research, and niche weighting to turn what you actually listen to into genuine compatibility signals.
How Niche Weighting Actually Works
The core idea behind Affinity Atlas matching: why shared niche interests are weighted more heavily than mainstream ones. The maths, the intuition, the research, and real examples across music, gaming, beer, and books.
Building a Transparent Algorithm
Every dating app has a matching system. None of them tell you how it works. Why Affinity Atlas shows its working - the black box problem, three levels of transparency, the match card interface, and what XAI research says about explainable recommendations.
Multi-Platform Identity Stitching
You are not your Spotify account. Or your Steam profile. Or your Untappd check-ins. How Affinity Atlas builds a unified compatibility profile from scattered data across platforms - with consent-first OAuth linking, data normalisation, and an honest look at the edge cases.
Privacy by Design in a Data-Hungry App
Dating apps collect more intimate data than almost any other category of software. How Affinity Atlas handles the tension between needing rich behavioural data for matching and respecting user privacy - with opt-in integrations, derived signals, and data minimisation.
Dating Apps Are a Public Health Issue
Sexual assault tracking suppressed by Match Group. Romance scam legislation in Congress. Cyberflashing criminalised in the UK. The evidence that dating apps are a public health issue is mounting - and regulators are finally starting to act.
The AI Girlfriend Problem
AI companion apps surged 700% between 2022 and 2025. They promise connection without vulnerability. But the research says they make loneliness worse - and they are reshaping what an entire generation expects from relationships.
The Elo Score Problem
Dating apps secretly rank you by desirability using a system borrowed from chess. The assumption: attraction is a zero-sum competition. Why that framing is both technically flawed and genuinely harmful - and what matching should measure instead.
Why "Designed to Be Deleted" Is a Lie
Hinge calls itself the dating app designed to be deleted. It is also owned by Match Group, a company whose entire business model depends on you not deleting it. The slogan, the financials, the features that contradict it, and what genuine deletion-readiness would look like.
The Loneliness Economy
Dating apps are a $6 billion industry built on the promise of connection. The research says they often make loneliness worse. The Surgeon General's advisory, the burnout numbers, the dopamine loops, and what connection-first design looks like.
Who Owns the Dating Market?
Match Group controls Tinder, Hinge, Match.com, OkCupid, and Plenty of Fish. The dating app market generated $6.18 billion in 2024. One company owns most of it. Here is the full picture - the consolidation, the revenue, the cracks, and the challengers.
The Hobby App Takeover
Gen Z is ditching dating apps for Strava run clubs, Goodreads book clubs, and Letterboxd film circles. The platforms replacing swiping are not dating apps at all - they are hobby apps where genuine connection happens as a side effect of shared passion.
What Niche Dating Apps Get Right (and Wrong)
From FarmersOnly to Feeld, vertical dating platforms promise community over volume. An honest survey of the landscape - six apps examined, what works, what does not, the hobby app phenomenon, and where Affinity Atlas fits.
Why Dating Apps Are Broken
The conflict of interest at the heart of modern dating apps. How Match Group, dark patterns, and subscription models keep you searching instead of connecting - and what a transparent alternative looks like.
The Data Architecture
How Affinity Atlas collects, structures, and standardises data from dozens of sources. Four data categories, hierarchical modelling, cross-domain interconnectedness, sparse data tiers, and the engineering challenges of cross-platform identity resolution.
How the Affinity Atlas Algorithm Works
The full deep-dive into niche weighting, Affinity scoring, engagement normalisation, and why shared obscure taste is worth more than shared mainstream taste.