Description
As artificial intelligence moves from research labs into hiring, lending, criminal justice, medicine, and transportation, its decisions increasingly shape human lives—often in ways its creators neither intended nor understand. This compelling exploration examines the alignment problem: the challenge of ensuring that machine-learning systems reflect human values, fairness, and judgment rather than reproducing hidden biases or pursuing flawed objectives. Through the work of pioneering researchers and the real-world consequences of algorithmic decision-making, the book considers questions of ethics, accountability, safety, and the future of human control. Engaging and deeply reported, it offers a timely account of the effort to make intelligent machines not only more capable, but more trustworthy.
Available Editions
hardcover
W. W. Norton & Company
Published Oct 6, 2020
356 pages
ISBN: 9780393635829
Reviews
No ratings yet. Be the first!
No written reviews yet.
Details
First Published
2020
Language
eng
Featured in a Challenge
Content Warnings
Content and trigger warnings for The Alignment Problem by Brian Christian.