Why I am Making a Privacy Focused AI Chatbot

When AI was first introduced, the internet was free to be scraped. Even though scraping was a legal grey area, it was done in order to create the first successful AI chatbots. We memed about how the internet was being sold back to us at a subscription cost. It was a closed and proprietary system feeding off of public data.

Before AI, no one made a post realizing it would make its way into a training set. No one posted a video thinking it would be used to train Sora. And when AI become successful, few questioned whether or not this was okay. Thankfully, artists have been the first to see a danger in AI, and they pull back their best work, but is that the internet we want?

With the advent of social media, we became privacy conscious, but social media became the spice of the internet, and we decided to overlook the fact that we did not have much control over our data. Who cares if Google can read all of our emails? Who cares if Facebook can read our DMs? Even when Mark Zuckerberg’s controversial remark became public, we continued to give him our data. "They trust me — dumb fucks.”

Years later, public email protocols are still being used, and now Facebook has turned Meta, hoping to collect much metadata and grab you by the eyeballs. To paraphrase the Joker, “No one bats an eye”.

I remember in high school, when weed was illegal, I showed a picture of a bong I had just purchased to a friend through a Facebook DM, and one day, the cops came to my door because my brother had gotten in trouble. Seeing me, the cop said, “I have a picture of you holding a bong”. Social services were called, and in my living room, the lady said, “Anything you post on the internet is there forever”. My trust in Facebook was destroyed. I could not believe that law enforcement somehow had access to my Facebook DMs. When Facebook locked me out of my account asking for a photo of my government ID to regain control, I deleted my account instead. Hide yo kids, hide yo wife.

We as westerners worry about social credit scores, but in reality, these systems are probably already being built secretly without the public being aware.

I think we have fallen asleep to privacy discussions we used to have back in the early days of social media. Now with the advent of AI, I feel that we need to have these discussions again.

One day AI will be ubiquitous. It’s too convenient not to be. We will all have access to Jarvis. Students will use it to accelerate their learning. Businesses will use it to build and serve. People will use it for journaling, telling it all their personal information, and it will be therapeutic, pointing out connections you were unable to make. It will manage your busy life, reminding you of upcoming events and appointments. Where is the irony here? Is it that AI meant to give us freedom will control us? Or is it that social media was meant to connect us, but we will all talk to AI instead?

In The Dark Knight, when Bruce Wayne reveals the mass surveillance system to Lucius Fox, who is played by Morgan Freeman, known for being the voice of God, he remarks, “This is wrong”. Bruce says it’s necessary to catch the Joker, reminding us of the argument that mass surveillance protects us. Lucius replies, “As long as this machine is at Wayne Enterprises, I won't be”.

Having tried AI, and seeing it’s potential for privacy abuse, I agree with Lucius’ sentiment, so I will be working toward creating privacy focused alternatives to proprietary AI chat systems. I hope you see the danger, too.

He who controls the spice controls the universe.