★★★☆☆ Honorable Mention
(Hard SF) An engineer checks out a complaint about problems with a home AI, but all the problems have plausible explanations. All but one. (1,666 words; Time: 05m)
This story is not available for free online; you have to purchase the issue to read it.
"The Explainer," by Ken Liu [bio] (edited by John Joseph Adams), appeared in Lightspeed Magazine issue 100, published on September 1, 2018.
Mini-Review (click to view--possible spoilers)
Review: 2018.502 (A Word for Authors)
Pro: This is an idea story, and it’s a great idea story.
In the real world, neural nets and other machine-learning systems learn to optimize what they’re told to, which isn’t always what the humans really want. In this case, the system has learned to optimize service calls. Bravo for authenticity!
I particularly liked the element that someone developed a technology to make it possible to debug neural nets.
Con: There isn’t much of a story here. No plot, no character development. The idea is great, but it’s all there is.
Other Reviews: Search Web, Browse Review Sites (Issue 100)
Ken Liu Info: Interviews, Websites, ISFDB, FreeSFOnline
Follow RSR on Twitter, Facebook, RSS, or E-mail.
Pro: This is an idea story, and it’s a great idea story.
In the real world, neural nets and other machine-learning systems learn to optimize what they’re told to, which isn’t always what the humans really want. In this case, the system has learned to optimize service calls. Bravo for authenticity!
I particularly liked the element that someone developed a technology to make it possible to debug neural nets.
Con: There isn’t much of a story here. No plot, no character development. The idea is great, but it’s all there is.
Other Reviews: Search Web, Browse Review Sites (Issue 100)
Ken Liu Info: Interviews, Websites, ISFDB, FreeSFOnline
Follow RSR on Twitter, Facebook, RSS, or E-mail.
No comments (may contain spoilers):
Post a Comment (comment policy)