AI helps add 10k more photos to OldNYC

Over the past two years I’ve quietly rebuilt major parts of the OldNYC photo viewer. The result: 10,000 additional historic photos on the map, more accurate locations, and a site that’s cheaper and easier to run—thanks to modern AI tools and the OpenStreetMap ecosystem.

OldNYC had about 39,000 photos in 2016. Today it has 49,000. Most of these changes happened in 2024, but I’m only writing about them now in 2026. Here are the three biggest improvements: better geolocation, dramatically improved OCR, and a switch to an open mapping stack.

1. Smarter Geolocation with GPT-4o

OldNYC works by geocoding historical descriptions—turning text like “Broad Street, south from Wall Street” into a latitude and longitude. Originally this mostly meant extracting cross streets from titles and sending them to the Google Maps Geocoding API. That worked well when the streets still existed—but many historical intersections don’t.

I now use the OpenAI API (gpt-4o) to extract locations from description text. For example, a photo of Public School 143 from 1930 could not be geocoded by title alone because the school no longer exists. However, GPT extracted the relevant intersections from the description. Using GPT located about 6,000 additional photos. Today OldNYC can locate roughly 87% of photos with usable location data, and about 96% of mapped images appear in the correct location.

2. Revolutionizing OCR

Most OldNYC photos include descriptions from the NYPL catalog—but on the NYPL site these are scanned typewriter images, not text. When I launched OldNYC in 2015, converting these images to text (OCR) was the hardest technical problem.

In 2024 I rebuilt the OCR system using gpt-4o-mini. The results were much better. Even in cases where the old OCR produced complete gibberish due to an unusual font, GPT transcribes it perfectly. Tools like OpenAI mean that OCR is a much easier problem in 2024 than it was in 2015.

3. Switching to an Open Mapping Stack

When OldNYC launched, Google Maps was the default choice. But over time, Google’s pricing model changed, making it expensive for a hobby project. I migrated the site to OpenStreetMap vector tiles and MapLibre.

The new stack has significant benefits. Beyond being cheaper, it allows for better customization. For example, I can remove anachronisms like highways and tunnels that didn’t exist in the 1930s, providing a more authentic historical experience.

The Future of OldNYC

There’s still plenty to improve. AI could extract additional information from images—identifying people, buildings, or indoor/outdoor scenes. I’ve also started contributing to OpenHistoricalMap. If it eventually includes full historical street grids for NYC, locating photos could become dramatically easier.