View all sketchnotes here.
“In mathematics and computer science, an algorithm ( i/ˈælɡərɪðəm/ AL-gə-ri-dhəm) is a self-contained step-by-step set of operations to be performed. Algorithms exist that perform calculation, data processing, and automated reasoning.”
More than ever, Algorithms are used to screen people for employment, to determine credit scores, to inform prison sentencing, and many other functions which shape peoples’ lives in important ways. But the assumptions and decisions which go into creating the algorithms can create outcomes that are not always in line with values like fairness, social mobility, racial impartiality, and so on. It’s important to not accept algorithms as a kind of truth machine, but rather to foster discussion about how to best use and shape algorithms to create the kind of outcomes that foster the kind of outcomes we want. For example, (to pick a simplified and hypothetical example) do we want an algorithm that rewards teachers for gaming the system to get a higher raise, or do we want an algorithm that results in students having better educational outcomes?
A panel of data experts led a discussion at Civic Hall yesterday about the necessity to interrogate algorithms–to question their goals, efficacy, and context in which they exist. Cathy O’Neil, Meredith Broussard, and Solon Barocas.
I captured the conversation with this set of sketchnotes. The event was sponsored by DataKind, The Microsoft Tech and Civic Engagement team, and was hosted at Civic Hall and was the first in a series of lunchtime events exploring data science called Machine Eatable.
Tech note: I created the sketchnotes in real time on an iPad using the Paper App.