Google Now Using ReCAPTCHA To Decode Street View Addresses
Have you started seeing images in online reCAPTCHAs that look suspiciously like house numbers pulled from Google Street View? Well, as it turns out, that's exactly what they are. Google confirmed it's currently running …
Context & Ripple Effects
Google's confirmation closes a loop it opened back in 2007, when it began blurring faces and license plates on Street View to defuse privacy complaints. House numbers were left readable — and they turn out to be the point: images shown in reCAPTCHA puzzles are genuine Street View crops, so every human solver is doing free transcription work that improves Google's address geocoding.
The story travelled unusually far for a single confirmation — pickups at InfoWorld, The Verge, Geek.com and Gizmodo UK all landed the same day — because it exposed reCAPTCHA, which publishers embed site-wide as spam defense, as an active data-collection pipeline for Google Maps rather than a neutral security widget.
First-order effects
- Every webmaster running reCAPTCHA is, knowingly or not, donating their visitors' solve time to decode Street View addresses for Google's mapping database.
Second-order effects
- Rival mapping and geocoding providers have no equivalent free-labeling channel, forcing them to pay for transcription labor or accept weaker address coverage against Google's zero-marginal-cost pipeline.
Third-order effects
- If the pattern holds, security friction itself becomes an asset class: whoever distributes the challenge widget controls a crowdsourced annotation workforce, raising the question of whether publishers will demand a share or defect to alternatives.
The trend: CAPTCHAs are evolving from pure spam filters into distributed human-computation systems that feed corporate datasets, with Google as the first mover at scale.