In practice, your devices are certainly gathering data about you as the user. However, it’s not exactly a matter of listening in. Rather, the apps you use and the data they collect are used in a form of machine learning, working to make predictions about what you’d be interested in and get that information in front of you. To do so, it considers a range of factors.
What is machine learning?
Put simply, machine learning is a type of artificial intelligence in which a computer system improves without being actively programmed. Rather than a programmer having to constantly add improvements, the program can use data to improve its own processes.
Experts typically split machine learning into two categories: Supervised and unsupervised learning. Real-life examples of supervised learning include classification, an integral part of data organization, while unsupervised learning focuses on more complex tasks. In the case of understanding how your phone gathers seemingly unknowable insights on you as a user and consumer, though, you’re likely more interested in a different set of data.
Where have you been lately?
In solving the mystery of how your phone seems able to read your mind, your location is a crucial clue. As a recent Twitter thread by privacy tech worker Robert G. Reeve describes, your phone can recognize the GPS location you’re in, and the devices that surround you there. In Reeve’s case, he spent a week at his mother’s home and used her chosen brand of toothpaste. After returning home, he found he was getting ads for that same toothpaste brand.
Your data can then be connected with the available data from the people around you. If it realizes you’re spending a lot of time with a smoker, for instance, and another friend who’s an avid couponer, you might find that you’re suddenly seeing ads for cheap tobacco.
When the algorithm is correct, you might follow that advertisement to purchase electronic cigarettes or pipe tobacco. Or, you may discuss that ad with your friends, in a striking example of this data influencing your behavior. They then might use that information in a number of ways, making a purchase based on your conversation or even finding that they, too, are now getting similar ads.
Which permissions have you approved?
It’s undeniable that many people agree to apps’ terms and conditions without reading or understanding them. More often than not, that agreement includes your consent to data sharing. Apple’s iOS recently implemented a privacy update in which apps ask permission before tracking your phone’s data.
While this App Tracking Transparency tool wasn’t released with the breadth expected, it’s nevertheless marked a change. Even as some developers try to circumvent these changes, users have noticed an influx of pop-ups asking for permission to track you—and the ability to opt-out, should you choose. Depending on your response to that prompt, you might have more data output than someone declining app tracking. Many apps have yet to implement ATT at all, further complicating users’ understanding of how much data a program has access to.
Who owns an app?
However, you often won‘t need to agree to certain sharing—different apps may access your data purely because they’re owned by the same company. Facebook, for instance, owns apps like Instagram and WhatsApp. Google owns YouTube and Android, amongst their plethora of G-branded services. Whenever you use one of these programs or services, the others have access to your data, so don’t be surprised if you start seeing Google ads for something you just saw on YouTube.
In short, your phone doesn’t need to read your mind or listen in on your conversations to target advertisements and other content to your interests and needs. Between types of machine learning, historical data, and the countless ways we input data into the algorithms, these devices know exactly what you’re looking for—often before you do.
