Two college students have used Meta smart glasses to create a device that can quickly identify any stranger walking by and reveal that person’s sensitive information, including home address and contact information, according to a demonstration video posted on Instagram. And while the creators say they have no plans to release the code for their project, the demo gives us a glimpse of a very possible human future—a future often found in dystopian sci-fi movies.
The two people running the project, AnhPhu Nguyen and Caine Ardayfio, are computer science students at Harvard who often post their technological experiments on social media, including 3D printed images and wearable missiles. But their latest test, first discovered by 404 Media, is likely to make a lot of people uncomfortable.
Instagram video Nguyen’s post describes how the two men built a system that feeds visual information from Meta Ray Ban smart glasses into facial recognition tools like Pimeyes, which scoured the entire web to identify where that person’s face appeared on the Internet. From there, a large language model takes a possible name and other information about that person. That name is then entered into various websites that may reveal that person’s home address, phone number, job or other organizational structure, and even the names of relatives.
“To use it, you just put on the glasses, and when you walk around people, the glasses can see if the person’s face is in the frame. This photo is used for personal analysis, and after a few seconds, their personal information appears on your phone,” Nguyen explains in an Instagram video.
Nguyen and Ardayfio call their project XRAY and it’s amazing how much information they’re able to extract in such a short amount of time. They are quick to point out that many of these tools have become widely available over the past few years. For example, Meta smart glasses with camera capabilities that look like normal glasses were released only last year. And the kind of LLM data release they’re getting has been happening for the last two years. Even the ability to look up incomplete social security numbers (thanks to all those data leaks you read about every day now) would only be possible at the consumer level starting in 2023.
As you can see in the video, they also talked to strangers and pretended to know those people from another place after looking up their information instantly.
“This program uses the LLM’s ability to understand, process, and integrate a large amount of information from various sources – to show the relationship between Internet sources, such as linking a word from one topic to another, and to logically convey the identity of a person and the details of a person in a text,” the creators said in the explanatory document posted. in Google Drive. “This collaboration between LLMs and reverse face search allows for automated and comprehensive data extraction that was previously not possible with traditional methods alone.”
The creators list the tools they used in their release, noting that anyone can request that those services remove their information. For face-reverse search engines, there are Pimeyes and Facecheck ID. Search engines that include personal information include FastPeopleSearch, CheckThem, and Instant Checkmate. As for social security number information, there is no way to remove that information, so readers recommend freezing your credit.
The students did not immediately respond to Gizmodo’s questions Wednesday morning. Meta also did not respond to a request for comment. We will update this post when we hear back. But for now, we should all be ready for this kind of technology to come out more widely as this kind of technology mash-up seems inevitable at the moment – especially if there are new glasses that are loved by the likes of Mark Zuckerberg. really become the norm.
It may take a long time for the big tech companies to get behind you, but as we’ve seen OpenAI is actually firing the first shot against AI consumers, any small startup can make this product happen and start one dominoes falling. large technology companies to start this future. Let’s cross our fingers and hope for the best, given the privacy implications. It sounds like no one will have any semblance of public anonymity once this ball gets rolling.