Technology and Society Book Reviews

Link to index page Link to privacy reviews Link to commerce, security, and the law reviews Link to culture reviews. Link to politics, security, and the law reviews Link to ethics, rhetoric, and metaphysics reviews. Link to science fiction reviews.

Title: Artificial Unintelligence

Author: Meredith Broussard

Publisher: MIT Press

Copyright: 2018

ISBN13: 978-0-262-03800-3

Length: 248

Price: $24.95

I purchased a copy of this book for personal use.

In Artificial Unintelligence, Professor Meredith Broussard of New York University cautions against technochauvanism, which she defines as the belief that technology in general (and computers in particular) can solve the world's problems. She acknowledges that technological advances have unquestionably made our lives better, but argues persuasively that technology is a human construct that reflects the biases and shortcomings of both our technical capabilities and human nature. Many technology problems are actually people problems, some of which are created or exacerbated by technology.

Versatile Yet Lacking

When hyping a product, pitchmen often ask "What can't it do?" In the case of technology, the answer is clearly "a lot". As Broussard notes, computers are very good at repetitive, well-defined tasks. Machine learning and artificial intelligence code uses iterative functions to implement techniques such as gradient descent or nearest-neighbor clustering to find optimal (or near-optimal) solutions and divide data points into meaningful groups. Those algorithms are based on data and, while it's true that more data leads to better decisions (to a point), it's also true that biases in the data are magnified in the resulting models. If crimes in certain neighborhoods are more likely to be reported, or if members of one ethnic group are more likely to be arrested, then algorithmic policing efforts will reflect those biases. Add in the human element, which affects everything from data collection to verification and interpretation, and you can identify many opportunities for unintelligence.

Textbooks, Self-Driving Cars, and Startup Buses

Before joining the faculty at NYU, Dr. Broussard was a features editor with the Philadelphia Inquirer. While she was there, her first-grade son brought home an assignment but had to leave the textbook at school. The teacher had promised to provide a code for online access, but never did so. When she spoke with school officials, they revealed that while there was a central data repository with information on the district's textbooks, it wasn't accurate or useful. After investigating the available data, which it turned out was spread across numerous spreadsheets, she discovered that there was an entire case of calculus AP exam prep books stored at a school when the staff thought they had none (the books were found in a locked closet). That instance was just one example of how teachers had to deal with incomplete or inaccurate information and use what Broussard termed an "underground economy" to get needed books when other schools in the system closed or their curricula changed.

Broussard also devotes two chapters to vehicle themes: self-driving cars and a startup bus. Her ride in a self-driving car in a nearly empty parking lot was harrowing. The car's software had a known bug, albeit one the developers thought they'd fixed, that caused it to accelerate dangerously toward a streetlight it should have avoided. The programmers, a team from Carnegie Mellon University, said "It happens" and said "This is high on my list of things I don't want to fix today" when someone opined which code module might be causing the problem.

Her other vehicle-related experience was on a startup bus. The idea was that individuals who joined a long-distance bus trip would be formed into teams and bash out a startup idea while on the way to the final presentation, where one of the final groups would be declared the winner and receive some small seed funding for their idea. Broussard's team created an app that let you calculate how much pizza you needed to buy for a party. It was coded in Node.js and ended up winning the competition, but she notes that she's not holding her breath for a big company to buy the technology. The reality, she notes, is that "software is rarely disruptive or innovative, and even more rarely is it both together." That's why entrepreneurs create a lot of businesses and are willing to fail fast...they need to throw a lot of ideas against the wall to see what sticks.

Healthy Skepticism

Professor Broussard teaches data journalism at NYU. The thorough, questioning approach inherent to that profession serves her well in writing Artificial Unintelligence. The journalistic ideal is to stand outside of the action so as to report facts and implications objectively, but her book is persuasive narrative non-fiction. I thought she struck a fine balance between personal experience and objective reporting—her opinions were clearly shaped by her experiences and she became part of the story, but she explicitly noted those caveats and her work benefited from those choices. Highly recommended.

 

Curtis Frye is the editor of Technology and Society Book Reviews. He is the author of more than 30 books, including Improspectives, his look at applying the principles of improv comedy to business and life. His list includes more than 20 books for Microsoft Press and O'Reilly Media; he has also created more than 50 online training courses for lynda.com. He received his undergraduate degree in political science from Syracuse University and his MBA from the University of Illinois. In addition to his writing, Curt is a keynote speaker and entertainer. You can find more information about him at www.curtisfrye.com and follow him as @curtisfrye on Twitter.

 

Link to new reviews.Link to list of reviews by publisherLink to page with contact information.