Say What You Mean, or When Not to Use Certain Technical Terms

The phase "Choose your words" are spelled out using wooden Scramble letter tiles on a white table. The word "your" is vertically spelled using the "o" and "r" in the horizontal words "choose" and "words", respectively.
Photo by Brett Jordan on Unsplash

Welcome to the first week of Daylight Savings Time for most of our readers in the US! Now that we are short one hour of sleep, it’s the best time to start with a thought experiment. The following is an excerpt from a recent library technology conference poster proposal:

“Patrons who visit an academic library with their smart devices (i.e., cell phones, laptops, tablets) connected to the campus Wi-Fi services would have their geolocation data, user ID, and time stamp stored in the Wi-Fi service provider’s system. The big data harvested provides a clear view of patron demographic information, including majors, classes being taken, along with other data… the use of Artificial Intelligence has helped the library to predict user behavior and thus be able to more closely tailor facilities, collections, and instruction to enhance student success.” (emphasis added)

Now for the question – what would you use instead of “Artificial Intelligence” in that excerpt? Take a moment to write down whatever comes to mind.

As we started exploring in our “On The Same Page” series, words are complicated. Sometimes they don’t let onto the complexity of the concept they represent, such as personal data. Other terms are prone to obscure, misdirect, or otherwise conceal the real-world consequences of the ideas and actions represented by those terms. Phrases like “artificial intelligence” and “machine learning” find widespread use in our lives without much thought into what they mean and the implications behind those terms in our understanding of technology. What can we use instead of these terms, though?

An excellent place to start is to say what you mean. The Center on Privacy & Technology at Georgetown recently announced that they will no longer use terms like “artificial intelligence” and “machine learning.” Instead, they will use the following guidelines to say what they mean:

1. Be as specific as possible about what the technology in question is and how it works.

2. Identify any obstacles to our own understanding of a technology that result from failures of corporate or government transparency.

3. Name the corporations responsible for creating and spreading the technological product.

4. Attribute agency to the human actors building and using the technology, never to the technology itself.

One example provided by the article takes the phrase “face recognition uses artificial intelligence” and replaces it with “tech companies use massive data sets to train algorithms to match images of human faces.” The latter phrase is specific as to what is all involved, including the human involvement behind the technologies referenced in the former term. The latter phrase also doesn’t conceal the process of facial recognition – it takes data from real human faces, and lots of it, to get an algorithm to determine a match of a face with an image correctly. But wait – where do the faces come from? What decisions are being made about which faces to feed into the algorithm? Do the people whose faces are being used to train this algorithm know that their faces are being used in training? What are the ultimate goals of the tech companies in creating this type of technology? Who are these tech companies in the first place?

Being specific about the technology, how it works, and the humans behind the technology better positions the readers in asking questions about the real-world impact of these technologies. It also attempts to make more apparent to the readers the potential harms that can come from these technologies, such as the potential of lack of consent from the people whose faces are being used for training and the potential bias in the data set itself based on who is included. Spelling out the specifics breaks us from using technical terms that we and our audience might not fully understand or be aware of the potential privacy risks and harms inherent in these technologies.

Let’s revisit the excerpt from the beginning of the post. With the Privacy Center’s guidelines in mind, what would you say instead of “Artificial Intelligence” in the last sentence?

(Bonus – Are there other technical terms in the excerpt that need to be spelled out? If so, what should be said instead of those terms?)

We invite you to share your answers with us! You can use the following form to share your answers with us. We are not collecting personal data such as IP address, name, or address for submissions. We’ll return to the exercise and share the responses in a future post, so stay tuned!