In the 1800s, saloons would entice clients to come in with an offer of free lunch, if they paid for a drink (Quoteinvestigator.com, 2016). This was the best-known start of the phrase “no such thing as a free lunch.” The poem “Somebody Pays” by Josephine Pollard (1887) says that somebody pays for the free things you receive.

Understanding AI personal data usage is essential to understanding your rights and privacy protection. Do you know what is collected about you with every software you touch? Should you check?

Let’s start with understanding basic AI vocabulary (without the marketing hype).

• Large Language Model (LLM) is when machines are trained to anticipate the next word or phrase, sounding human. To do that, they need to gather huge amounts of words and sentences from the internet (University Libraries FAQ) . In conventional programming, humans create step-by-step directions including checks on certain situations (if-then-else). LLM is where the machines absorb examples and begin to predict the answers. The key to success with LLMs is to provide enough data and processing capabilities.
• Artificial Intelligence (AI) is built on LLM using historical data to make predictions.
• Generative AI creates new outputs – words, images, music, and art.

Where does AI get the information needed to learn and respond better? From us. Systems have been collecting data on us for years, but the scale of data needed with AI is much larger (King & Meinhardt, 2024). AI is less transparent and, in general, we have been trained not to look or work at protecting our data as much.

For example, a colleague was suggesting some AI systems to try with our students. One seemed intriguing, so I decided to test it out. First, I need to at least glance at the privacy policy. I am not a lawyer and in general, the privacy policies can be hard to follow. I always check first to see how they address students under 18 years old, even if I’m using it for myself. Measuring protections for children is a good way to see how a company treats its patrons.

Explainpaper (“Understand complex research papers in minutes with AI explanations”) has no information about its company or privacy policies on its main page. No information about who owns it or a board of directors, etc. I have not actually signed up to see if anything appears. I can find nothing using general searches about the company. I tried searching “Explainpaper CEO” and found two links in LinkedIn. I sent the CEO a private message over a month ago and have not received anything.

Do I think Explainpaper is collecting your data? Absolutely. Most systems collect computer information (IP, location, amount of usage, other programs in use, etc.) and some personal information. What do they actually collect? I don’t know and neither do you. Their concept is seductive, especially to a student trying to get through a large number of papers quickly. (Note: I teach students to read the abstracts first.)

Would I use it? No. Its utter lack of transparency is scary. Who is holding the reins and why are they hidden? What are they doing with your data?

King and Meinhardt (2024) recommend changing policies to require people to opt-in to data collection instead of opt-out (if that is even an option). They also recommend identifying and sharing how data is moved through the various companies and protecting it better. Unfortunately, these are plans for the future. Right now, we need to be aware and ask questions. If not for ourselves, then we need to help protect our students.

References:

King, J. & Meinhardt, C. (February 22, 2024). Rethinking Privacy in the AI Era: Policy Provocations for a Data-Centric World. Stanford University Human-Centered Artificial Intelligence (online). Accessed December 30, 2024. https://hai.stanford.edu/white-paper-rethinking-privacy-ai-era-policy-provocations-data-centric-world

Pollard, J. (1887). The Montgomery monitor. (Mt. Vernon, Montgomery County, Ga.) 1886-current, June 01, 1887, Image 1. Digital Library of Georgia, a part of GALILEO. Accessed December 30, 2024. https://gahistoricnewspapers.galileo.usg.edu/lccn/sn89053231/1887-06-01/ed-1/seq-1/ocr/

Quoteinvestigator.com (Aug. 27, 2016). Quote Origin: There Ain’t No Such Thing as a Free Lunch — TANSTAAFL. Accessed December 30, 2024. https://quoteinvestigator.com/2016/08/27/free-lunch/

University Libraries FAQ. What is a large language model? University of Arizona. Accessed December 30, 2024. https://ask.library.arizona.edu/faq/407985