Editing Avarice Engineer

The strength I leaned on in my career was avarice engineering. I mostly worked for advertising, marketing, or retail companies. My job was to increase revenue. And I was on the leading edge of altering people's desires, algorithmically.

At its core, it is simple. Look at every behavior you have seen from a person, quantify that data into an embedding - a list of floating point numbers, or vector of scalars - and based on that understanding of a person and how they react, you put the stimuli in front of them that will make them do what you want.

In 2008, for a few months, I may have been the best in the world.

I created and deployed a new method of ad targeting that intentionally used machine learning to manipulate people's emotions to increase their spending. I tripled sales compared to our cube farm of demographic targeting experts.

For years, I dined out on the following story:

We could, right now, create an algorithm that looks at your phone's historic location on Google Maps, and determine whether you attend alcoholics anonymous meetings. I can infer when you have a moment of weakness and miss a few meetings.

There already exist, deployed as clients and built-in to many services, systems that monitor people's emotional states in real time, based on their social media postings. Those two identities are tied to each other in various ways, most easily because your cell phone just tells everyone your unique ID, across domain boundaries.

I can know when you have had a weakness for drink, and what was the sequence of your emotional states leading up to that moment. I can look at your browsing history and see what stories you read in Google News that day. I can act on that distilled knowledge. I can surface the Gaza story, or the Ukraine story, or the Noem story. I have dozens of versions of that story and I know the emotional and motivational dimensions of each with dozens of degrees of freedom.