Design X AI - 03
ALGORITHMS
If we are to consider what is new and different about AI, we should acknowledge what is not new about it, and how it is to be used and implemented at scale. The key foundation of Artificial Intelligence is the almighty algorithm. To speak in layperson, not computer-science terms whatsoever, algorithms are the building blocks of advanced AI models. A bundle of algorithms are used to calculate results while processing data sets, and the results are used to train the AI model to make predictions, and guide its decision making.
The algorithm is the foundation of most computer science. It is the mathematical function which turns an action into a result, or a sea of data points into useful patterns of information. These algorithms have been used by organizations, corporations, and businesses to help them achieve their goals and create new value, without the need for human error or involvement for decades.
Here are some delightful examples of how algorithms and large data sets help create value: Insurance companies track user behaviors to determine how likely they are to get in a car accident or otherwise cost the insurance company. How you are defined by their algorithm determines your insurance rate and service. But a car insurance company doesn’t only track your driving, they keep data on everywhere you go, where you spend your money, your smartwatch data, everything you sign away in the User Terms and Conditions, and you might get a rate significantly higher than someone with a worse driving record because their algorithm has determined that you are more likely to be involved in an accident due to any number of other abstract factors.
Algorithms are used to create each person’s individual home feed of content on Instagram, Facebook, X to maximize engagement and generate increasingly specific targeted advertisements. This basic function of an algorithm in social media is based on the collection and processing of an individual's digital footprints within their data set to mold behavioral trends toward an ideal mode of the platform’s benefit. A baker at the bakery will notice what bread is purchased and which gets left behind, and change what he offers in hopes that he will gain more customers and sell more bread to each customer that comes in. It is in the nature of business to seek growth in this way. Similarly, it is Meta’s goal to get more users on their sites, to spend more time on their sites, and to sell advertisements with an ideal ad-to-user fit.
They are achieving what the algorithm has set out to do, but at what cost? It is not incentivized for truth, balance, or social harmony. The outcome being pursued is more engagement, and some of the carrots and sticks which reliably lead to this across people groups are envy, crisis, and outrage. See Cambridge Analytica, Rohingya Genocide.
Are you annoyed, or creeped out by how accurate and timely targeted ads and content can seem? Consider that this is possible because 1. You fit into their data set, and their algorithm is predicting your behavior very well. 2. This is what gets sold to you, but the same data profile is used by a plethora of other organizations you depend on.
The digital-age is defined by an algorithmic information network which has mapped people as data points. We ask the algorithm for a specific outcome, and it delivers. We don’t need to look under the hood and dig into all the little details, because the algorithm is doing the thinking for us! Well look around, from mass misinformation, to social isolation, increasing inequality, and political division, if the algorithms driving our world were optimized for making it a better place, they’re failing. But as the saying goes, the system is not broken, it works perfectly to do what it was built for.
———————————————
DESIGN X AI is an 8 part blog series written 100% by my fallible human brain. I’m a designer, not an AI expert.