from the I-always-liked-that-R.-Daneel-Olivaw-fella dept.
Research finds trust in algorithmic advice from computers can blind us to mistakes:
With autocorrect and auto-generated email responses, algorithms offer plenty of assistance to help people express themselves.
But new research from the University of Georgia shows people who rely on computer algorithms for assistance with language-related, creative tasks didn't improve their performance and were more likely to trust low-quality advice.
[...] The paper is the second in the team's investigation into individual trust in advice generated by algorithms. In an April 2021 paper, the team found people were more reliant on algorithmic advice in counting tasks than on advice purportedly given by other participants.
This study aimed to test if people deferred to a computer's advice when tackling more creative and language-dependent tasks. The team found participants were 92.3% more likely to use advice attributed to an algorithm than to take advice attributed to people.
"This task did not require the same type of thinking (as the counting task in the prior study) but in fact we saw the same biases," Schecter said. "They were still going to use the algorithm's answer and feel good about it, even though it's not helping them do any better."
[...] Schechter and colleagues call this tendency to accept computer-generated advice without an eye to its quality as automation bias. Understanding how and why human decision-makers defer to machine learning software to solve problems is an important part of understanding what could go wrong in modern workplaces and how to remedy it.
"Often when we're talking about whether we can allow algorithms to make decisions, having a person in the loop is given as the solution to preventing mistakes or bad outcomes," Schecter said. "But that can't be the solution if people are more likely than not to defer to what the algorithm advises."
Journal Reference:
Bogert, E., Lauharatanahirun, N. & Schecter, A. Human preferences toward algorithmic advice in a word association task [open]. Sci Rep 12, 14501 (2022). DOI: 10.1038/s41598-022-18638-2
(Score: 4, Funny) by SomeRandomGeek on Tuesday September 27 2022, @10:12PM (4 children)
My spell checker insists this entire article is bullshit.
(Score: 3, Funny) by DBCubix on Tuesday September 27 2022, @10:28PM (2 children)
Clippy says that's a lie!
(Score: 4, Interesting) by looorg on Tuesday September 27 2022, @10:47PM (1 child)
Just looking at the paper I am wondering if they asked Clippy for advice when they wrote it. It sure does look like they asked and took some of that low quality algorithmic advice themselves and it turned out it really was shit. So theory confirmed!
(Score: 4, Funny) by Runaway1956 on Tuesday September 27 2022, @11:21PM
If three people with really small numbers say the article is shit, there's no reason for me to read it!
Abortion is the number one killed of children in the United States.
(Score: 0) by Anonymous Coward on Tuesday September 27 2022, @11:20PM
> My spell checker...
Well, this is the red site. Are you sure you didn't mistake text color for the SN banners? (grin)
(Score: 5, Insightful) by darkfeline on Wednesday September 28 2022, @02:34AM
It's like what Joel describes in his classic article
https://www.joelonsoftware.com/2002/11/11/the-law-of-leaky-abstractions/ [joelonsoftware.com]
New tools are not a substitute for understanding how things work.
"Why teach math when you'll use calcuators as an adult."
This. This is why. If you don't know what's correct, you don't know whether your tool is functioning correctly or not.
Join the SDF Public Access UNIX System today!
(Score: 2) by lentilla on Wednesday September 28 2022, @02:47AM (1 child)
Having largely managed to avoid Microsoft Office for the last decade, I have recently had the experience of writing emails where the software critiques my writing. It's interesting. Sometimes it makes good suggestions. Other times I laugh and think, no, that's not how I want to communicate. As darkfeline [soylentnews.org] opines above "New tools are not a substitute for understanding how things work".
Years ago I remember Wordperfect 5.1 had a grammar checker. By today's standards it was probably pretty basic, but I did enjoy seeing the statistics. To this day, I haven't found a replacement. (By replacement, I mean a) something that runs locally; b) GPL or equivalent licence; and, c) is a command-line utility that eats text.) I remain flummoxed - you would have thought some Nerd Of Linguistics would have written something as a pet project. Any suggestions?
(Score: 2) by Reziac on Wednesday September 28 2022, @04:26AM
WP5.1DOS did not come with a grammar checker. But there were add-ons that would do the job; I expect that's what you're thinking of. (I think WP5.2Win came with one, but I never did more than a passing fling through it so don't recall. But I still have WP5.1 installed on yonder DOS PC.)
As to suggestions, I got nothin'. I do my own grammar. :)
And there is no Alkibiades to come back and save us from ourselves.
(Score: 2) by Freeman on Wednesday September 28 2022, @01:45PM
I would be happier, if autocorrect was less stupid. Instead of helping me enter an incorrect word. Since, it obviously knew what I was thinking.
Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"