May 3, 2024: Double Feature–Fine-Grained Air Quality Sensing with Internet-of-Things & Soothsayers, Illusionists, Con Artists, and ‘Artificial Intelligence’

In our second 2024 SIGCAS WIP talk, we will have two presenters:

Julia Gersey will give a short talk on “Fine-Grained Air Quality Sensing with Internet-of-Things”
Dr. Trystan Goetze will give a longer talk on “Soothsayers, Illusionists, Con Artists, and “Artificial Intelligence”
At the end of the session, we will have approximately 15 minutes of shared QA with the speakers
Fine-Grained Air Quality Sensing with Internet-of-Things

Particulate Matter, which is 2.5 micrometers in thickness (PM 2.5), contributes to premature death, heart attack, decreased lung function, and other health effects. Yet, it is in the air we breathe every day. For Cuyahoga County, an area of 459.07 square miles and a population of 1,245,337 (as of 2020), only one air quality (AQ) sensor monitored PM 2.5 levels at the start of this research project. In our initial deployments, we use Raspberry Pi Models with a Plantower PMS5003 particulate matter sensor to capture PM 2.5 from the sensors and send the data to our web service as a low-cost and low-power approach. Besides increasing the number of sensors that report AQ data to the public, we have partnered with a local school to create and teach a Computer Science and Engineering curriculum to allow students to build and deploy their sensors within their community.

Our current work involves creating our public-facing website/mobile applications, teaching our curriculum at a local middle school, and expanding our deployments across Cleveland. Future work will deploy additional sensors in partnership with PCs for People and analyze data collected over extended periods. In addition, we will partner with Slavic Village Development to meet with the community to better understand how community members perceive air quality and how it impacts their daily lives. While the sensors we deployed may not be as accurate as regulatory sensors, previous research has demonstrated that low-cost sensors’ accuracy is sufficient to provide general trends.

About Julia

Soothsayers, Illusionists, Con Artists, and “Artificial Intelligence”

In an obituary for James Randi, the illusionist and debunker of psychic frauds, physicist Philip Ball remarks that “advances in science don’t banish credulity, but create new stages for it.” No such scientific advances illustrate this potential for deception quite like the recent public fascination with artificial intelligence. Doubtless, the achievements of machine learning in the last few years are impressive. However, researchers continue to show that claims of A.I. applications’ usefulness are routinely overstated, while harms of poor A.I. design or deployment are significant. Part of the problem is that A.I. developers have engaged in a dishonest kind of conceptual engineering, misdirecting how the public thinks about A.I. At the same time, both the conceptions of A.I. that developers peddle to the public and those which developers themselves seem to hold are continuous with the deep history of automata and divination, practices which rely either on illusion or deception. This paper examines this conceptual history, from ancient automata to the Mechanical Turk, from the Turing test to the present day. In view of the fundamental connection between A.I. and magic, the paper then comments on the ethics of illusion. As Randi himself might say, not all illusion amounts to unethical deception. But the same tricks which offer harmless amusement or genuine insight may be used to manipulate and exploit those who misunderstand the nature of the trick.

More about Cornell’s Bovay Program in the History and Ethics of Professional Engineering

More about Dr. Goetze
Registration is Free, but required