SoylentNews is people
SoylentNews
posted by
martyb
on Thursday December 17 2015, @10:20AM
from the better-improve-your-penmanship dept.
from the better-improve-your-penmanship dept.
Sanskrit, Tibetan, Gujarati, and Glagolitic were among 50 handwritten languages researchers used to test a computer program that proved to be as good, or better, than humans at recognizing the figures – a cognitive step for machines, and a leap forward for the potential that coders could build more sophisticated Artificial Intelligence (AI) in the future.
The program, developed by three researchers whose findings were published last week in Science, can recognize handwritten drawings after only viewing the figures a few times and also passed a basic Turing test.
Bad news for outsourcing centers in Gujarat. Hyderabad still safe.
This discussion has been archived.
No new comments can be posted.
Artificial Intelligence Passes the Turing Test of Penmanship
|
Log In/Create an Account
| Top
| 8 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Not recommended for children.
(Score: 0) by Anonymous Coward on Thursday December 17 2015, @10:42AM
For a Turing test, you "talk" to the computer. That is, the computer has to understand (or at least be able to pretend to understand) what you write. Recognition of hand-written characters is not even a requirement for that.
Don't get me wrong, this is certainly a great achievement. But it is sold as something it isn't: Passing the Turing test.
(Score: 3, Insightful) by wonkey_monkey on Thursday December 17 2015, @10:46AM
On a Turing test – a thought experiment devised in 1950 by British computer scientist Alan Turing to compare a machine's ability to think to that of a human – BPL consistently performed well. People could not tell the difference between figures drawn by the computer or human participants.
So there was a "can humans spot the computer" element, but it was about what the computer output, rather than how it responded to input.
systemd is Roko's Basilisk
(Score: 1, Insightful) by Anonymous Coward on Thursday December 17 2015, @11:37AM
It gets worrying once computers can solve unexpected problems. Until then they are mere servants and we should be worried more about people in power than their machines.
(Score: 1) by rigrig on Thursday December 17 2015, @01:42PM
How can a program be better than humans at recognizing figures, isn't the goal to get the result a human would pick?
No one remembers the singer.
(Score: 2) by hendrikboom on Thursday December 17 2015, @03:01PM
The real goal is to pick the what the human intended when he wrote it. Some people have trouble doing that with their own handwriting.
(Score: 2) by tibman on Thursday December 17 2015, @03:07PM
Better than the human average maybe?
SN won't survive on lurkers alone. Write comments.
(Score: 2) by wonkey_monkey on Thursday December 17 2015, @03:46PM
The goal is to recognise the symbols correctly. Humans can't always do it (especially, but not only, if they are unfamiliar with the written language presented), so there's room for improvement.
systemd is Roko's Basilisk
(Score: 3, Funny) by Zinho on Thursday December 17 2015, @03:19PM
So much for that method of keeping bots out of our forums.
Of course, we're approaching this threshold from the other direction as well; there may well be a day that Neil Harbisson [wikipedia.org] and Steve Mann [wikipedia.org] have an existential crisis when signing up for web forums. [thepunchlineismachismo.com]
The ADA in America would probably protect the rights of people who become cyborgs due to restorative prosthetics [wikipedia.org], but voluntarily enhanced humans may experience discrimination in the future. Early adopters, like body modders, probably know what they're getting into these days; it'll be interesting to see how attitudes shift in response to increased uptake of cybernetic technology.
"Space Exploration is not endless circles in low earth orbit." -Buzz Aldrin