Blog Post #8: Living in the Age of AI

When I’ve talked about The Algorithm™, it's usually in light reference to curated internet feeds. It’s not unusual for people to joke about its aggressive effectiveness. But returning to my computer science roots, I want to break down why it becomes genuinely terrifying when paired with modern AI.

An algorithm is “a computerized process that evaluates data to perform actions … without case-by-case human oversight”. Commonly, this looks like YouTube recommending related videos: ‘if you liked that, then maybe you’ll like this’ but automated. And honestly, when I first noticed how personalized my advertisements were, I wasn’t bothered by it. Humans simply can't process data at the scale machines can and the internet contains an overwhelming amount. So yes, I appreciated the help shifting through it all.

However, algorithms were never innocent. The breach of privacy was already real: a casual in person conversation about holes in my boots instantly turned into targeted ads for winter boots nearby. And the inclusion of AI takes standard surveillance to psychological profiling.

Training data can include highly personal information, such as menstrual cycles, credit card transactions, and even how confidently you type or whether you keep your cell phone battery charged. That data doesn’t remain isolated— it’s used to create a detailed psychological profile utilized by an apathetic AI equipped with decades of persuasion research—and built to exploit.

For example, based on my profile, the algorithm recognizes that while I doom scroll between 2PM and 5PM, I rarely buy anything during the day. Late at night, however, I'm exhausted and far more likely to impulse buy something I saw earlier on Amazon.

The AI solves this "complex problem" by analyzing which video snippets caught my attention during the afternoon. It plants the seed by showing me those products at a high price throughout the day. It knows I’ll ignore the ad initially, but it successfully hooks me on the product. Later that night, it hits me with the exact same product at a ‘limited discount’, manufacturing urgency and preying on my bad habits.

But the consequences go far beyond just advertising. It’s not a surprise to me that companies have conducted experiments to see if they could nudge real-world behavior, such as voting, or even manipulating users to feel "happy or sad" by adjusting their feeds. And when the information put into an AI model already contains historical biases, the resulting algorithms can reinforce or even amplify prejudice in extremely important areas like criminal justice and healthcare.

The scariest part isn't that a machine seems to know us better than we know ourselves— it’s that at the end of the day, an algorithm is still just a tool. The real horror is how the power of that tool is being used to secretly dictate how we act.