These days, Ki sometimes felt his job was the flip side of that craze. As a social media specialist with the Force, he spent long hours creeping through a public gallery of exhibitionistic Facebook, Instagram, Chirp and ClickChat users. Literally and figuratively barging through their open digital doors while they were naked. The accounts were vast, fertile and largely unguarded fields of useful information waiting to be harvested. But for all the time he’d put in over the past few months and all his legitimate reasons for poking around in people’s lives, the job still felt uncomfortable and voyeuristic. He didn’t know what compelled them to expose themselves to the whole wide world. He didn’t get how they could be so reckless putting things out there for anyone to see...until they were hurt or embarrassed by it.

At his workstation on the Terminal’s third floor Cyberdogs division, Ki had started out digging up info about Stephen Gelfland with a routine, methodical set of procedures, entering Gelfland’s biographical info into his search tool, then stealthily tracking mentions of his name across a plethora of different online sources—employment agencies, public records, LinkedIn, e-sellers, online dating sites and, of course, social media platforms.

It turned out Gelfland was a very private individual. His only SMP was a Facebook account that had been inactive for years. He therefore didn’t have many online contacts.

Still, Ki managed to get hold of a few web images. One was about a decade old and scraped from the State University of New York Maritime College’s Eight Bells yearbook, which was archived in the school’s digital library. Another was a much more recent US State Department passport photo. Since passports were public record, Ki hadn’t needed court permission to pull it.

Ki’s next step was uploading the pics to his Mirror Image face recognition app. The AI-powered search engine found and compared visually similar images of people using facial geometry, converting it into mathematical algorithms. It then measured how alike they were and served out the best matches. Mirror Image was intelligent enough to compensate for the filters people used when they posted photos on the internet, providing direct links to the sites where it found them.