/
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

Google details how it is using AI to improve search, including indexing individual passages from webpages, allowing searches for songs by humming, and more

Kyle Wiggers / VentureBeat :

VentureBeat Kyle Wiggers

Context & Ripple Effects

This October 2020 disclosure reads differently after six years of follow-on coverage: the two headline features — passage-level indexing and humming-based song search — were early moves away from matching whole documents toward understanding queries and content at a finer grain. A year later Google layered its MUM model onto Search with features like “Things to know”, extending the same pattern from retrieval to multi-step comprehension.

The trajectory since has been consistent: multisearch went global in 2023, Lens gained voice-driven video search in 2024, and by 2025 the Web Guide experiment was grouping results into query aspects automatically. That arc makes this 2020 piece the baseline for measuring how much of Search Google has since rebuilt around AI.

First-order effects

  • Passage indexing changes which pages win long-tail queries — individual paragraphs on otherwise obscure pages can now rank against full documents, directly altering traffic distribution for publishers and SEO practitioners.

Second-order effects

  • Every input modality Google adds after this — camera, voice, humming, then multisearch and Lens video search in later coverage — reduces dependence on typed keywords, forcing site owners to optimize for intent and passages rather than exact-match terms.

Third-order effects

  • If retrieval keeps shifting toward AI interpretation of content, the endpoint already visible in the corpus is generation: Google's experiment replacing news headlines in results with AI-written ones shows the platform moving from surfacing publishers' text to producing its own — with publisher exposure to that substitution as the unresolved structural question.

The trend: Google Search is being progressively re-architected from ranking webpages to interpreting and generating answers across text, image, and audio — a shift whose groundwork this 2020 passage-indexing and hum-to-search announcement laid.

Discussion

  • @starfire2258 Sean Hollister on x
    This is what I wanted Shazam to be from day one. You can whistle, too! So cool. Needs some work, though. It managed to recognize the Für Elise, but not Beethoven's 5th, and had a lot of trouble with Eine Kleine Nachtmusik. https://www.theverge.com/...
  • @tsuvik Vikas SN on x
    Tried Google's ‘Hum to search’ with a Hindi song and it worked quite good https://blog.google/...
  • @epro Emil Protalinski on x
    “Lastly, Google says it will let users search for songs by simply humming or whistling melodies, initially in English on iOS and in more than 20 languages on Android.” The commercial for this writes itself. https://venturebeat.com/...
  • @baxiabhishek Abhishek Baxi on x
    Now you can search songs on Google by just humming, and I think it is mind-boggling. - “Hey Google, what's this song?” (Or use Google Search app) - Hum for 10-15 seconds - And you'll get the result. It really works! For Hindi songs too. https://blog.google/...
  • @daveleeft Dave Lee on x
    Surely Google's biggest innovation since web search https://blog.google/...
  • @google @google on x
    Do you know that song that goes, “da na na na na do do?” We bet Google Search does. 😉 Next time a song is stuck in your head, just #HumToSearch into the Google app and we'll identify the song. Perfect pitch not required → https://blog.google/... #SearchOn https://twitter.com/...
  • @kayoberbeck Kay Oberbeck on x
    Yeah, THAT'S exactly what I needed! Song stuck in your head? Just hum to search @google https://blog.google/...
  • @itskyleharris Kyle Harris on x
    After years of google secretly listening to us hum songs poorly and never remembering song titles, they finally released #HumToSearch https://twitter.com/...
  • @salgar Satyajeet Salgar on x
    Really excited to get this out in the world: Hum to Search.. So many songs stuck in my head as part of testing this. :-) https://blog.google/...
  • @searchliaison @searchliaison on x
    Don't know the name of that song stuck in your head? Just hum it to Google Search. Really. We'll look for matches: https://blog.google/...
  • @searchliaison @searchliaison on x
    We shared a lot of new innovations in Search today as part of our #SearchOn event. Here's an overview covering things like spelling improvements, understanding passages and video moments, AR & Lens in Search and “hum to search” to find songs: https://blog.google/...
  • @thekenyeung Ken Yeung on x
    Here's what Google announced, in case you'd like a summary. https://venturebeat.com/...