/
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

Twitter says it can now figure out which tweets users find funny based on what they like and retweet and can show them personalized funny tweets

Karissa Bell / Engadget : Tweets: @isabellaturch , @lanceulanoff , @jtlol , @db , and @joshuaogundu Tweets: Isabella Turchetta / @isabellaturch : We also know that humor ranges, so we'll show you a lot of different Tweets starting out. Let us know what you like by liking/RTing the ones you enjoy, and selecting Not Interested on Tweets you don't. We'll then show you more Tweets that match what you like. Lance Ulanoff / @lanceulanoff : I will find my own funny, thank you very much https://www.engadget.com/... @jtlol : Or you could leave me alone https://twitter.com/... Himbo Depot / @db : This is likely just the beginning. Imagine this rolled out to any category you can think of. The future is now. https://twitter.com/... Josh / @joshuaogundu : Whoever is leading product at Twitter now is making sure they ship https://twitter.com/...

Engadget Karissa Bell

Context & Ripple Effects

Twitter's personalization push has been building for years: personalized news alerts based on followed accounts arrived in 2018, and 2020 brought a wave of conversational controls like reply limiting. On the same day as this humor feature, Twitter was also testing "humanization prompts" that surface shared interests and mutual followers in replies.

The new capability extends the same behavioral-signal playbook from topical interests to subjective taste: likes and retweets become training data for what a user finds funny, with Twitter explicitly asking users to like, retweet, or mark tweets Not Interested to tune the model. The user reaction quoted in the piece — resistance to being algorithmically assigned a sense of humor — foreshadows the timeline-control debates that later surfaced when Twitter for web stopped defaulting to the algorithmic feed.

First-order effects

  • Users' likes and retweets now feed a humor-inference model that reshapes what Twitter shows them, with an explicit feedback loop (Not Interested) for correction.
  • Twitter's own product team is soliciting engagement as training signal, making casual interactions — a like, a retweet — carry curation consequences beyond visibility of the original tweet.

Second-order effects

  • The feature deepens the algorithmic timeline's pull against the chronological one, sharpening the user split that Twitter later acknowledged by letting the web client remember which timeline was last open.
  • More inferred-taste surfaces mean more targeted engagement inventory, strengthening the ad-relevance case for the same behavioral data that personalized news alerts were built on.

Third-order effects

  • If platforms keep inferring ever-more-subjective preferences — humor here, shared interests in the humanization-prompt tests — algorithmic curation becomes the default interface for taste itself, raising the stakes of transparency moves like Twitter's 2023 partial open-sourcing of its algorithm.
  • The backlash pattern visible in the quoted user reactions points toward persistent demand for manual controls, pushing platforms into a two-track product: inferred feeds by default, opt-out chronology for those who resist.

The trend: Social platforms are extending behavioral inference from what users care about to what they find funny, making taste itself a machine-learned product surface.

Discussion

  • @isabellaturch Isabella Turchetta on x
    We also know that humor ranges, so we'll show you a lot of different Tweets starting out. Let us know what you like by liking/RTing the ones you enjoy, and selecting Not Interested on Tweets you don't. We'll then show you more Tweets that match what you like.
  • @lanceulanoff Lance Ulanoff on x
    I will find my own funny, thank you very much https://www.engadget.com/...
  • @jtlol @jtlol on x
    Or you could leave me alone https://twitter.com/...
  • @joshuaogundu Josh on x
    Whoever is leading product at Twitter now is making sure they ship https://twitter.com/...
  • @willoremus Will Oremus on x
    I wonder if Twitter's AI predicted that I would find it funny that a company that can't tell who's a Nazi or a pornbot claims to have solved humor https://twitter.com/...
  • @db Himbo Depot on x
    This is likely just the beginning. Imagine this rolled out to any category you can think of. The future is now. https://twitter.com/...
  • @hradzka David Hines on x
    “But I don't understand. How did Twitter suddenly start boosting racist shitposters?” “Uh, well —” https://twitter.com/...
  • @masonpelt Mason Pelt on x
    I only retweet things ironically. https://twitter.com/...
  • @isabellaturch Isabella Turchetta on x
    To power Topics like Funny Tweets, we need to understand the emotions behind the Tweets. So for the first time, we're using an ML model that understands nuanced emotions by learning from Emojis to get a broader understanding of what people think is funny.
  • @isabellaturch Isabella Turchetta on x
    Screenshots of funny Tweets are shared online all the time, and when people come to Twitter to get in on the fun, these Tweets should be easy for them to find. So, we launched Funny Tweets so you can catch the highlights of what people are laughing at.
  • @stevekovach Steve Kovach on x
    Good news for me and my hilarious tweets https://twitter.com/...
  • @caseynewton Casey Newton on x
    Twitter has ]announced more product launches and changes today than it did in all of 2014. https://twitter.com/...