When it comes to making slavery’s archive accessible, artificial intelligence has enormous potential.

Here’s why.

Right now, turning information from archival records into searchable data is an extremely labor-intensive task, because the work must be done manually, one record transcription at a time. As a result, most databases containing information about enslaved ancestors take many years to build.

In theory, however, there is a tool that could do a lot of this work for us. It’s called optical character recognition (OCR).

Optical Character Recognition (OCR) is a technology that converts images of text (like scanned documents or photos) into machine-readable, editable data (text).

Unfortunately, OCR is not yet sophisticated enough to accurately convert images of most historic records, including primary sources from slavery, into textual data.

However, with the help of artificial intelligence, OCR can be trained to effectively and efficiently encode historic documents as textual data.

This improvement in OCR’s capabilities would drastically reduce the time and effort required to produce data from archival records.

So, how do we do this?

Machine learning (ML) is a subset of artificial intelligence that enables computers to learn from existing data what they need to know in order to “understand” (encode) new data.

By training a machine learning algorithm (a set of rules or instructions a computer follows to learn from data) on existing databases built from primary sources, we can, therefore, “teach” computers how to convert other primary sources into textual data.

In order for this to work, however, we must train our algorithm on a lot of high-quality, existing data.

Through its partnerships with Kinfolkology (the home of the Oceans of Kinfolk database) and Freedom on the Move, the Turner-Hines-Franklin Institute will train a machine learning algorithm on the two largest databases of enslaved ancestors in the world.

In this way, AI will play a critical role in the Turner-Hines-Franklin Institute’s efforts to defragment slavery’s global archive.

Artificial intelligence has enormous promise, but there are dangers to consider, too.

On the Perils of Artificial Intelligence in Slavery’s Archive