Live data from Hacker News

A jackal's nose for the job

rbth.com

1–10 of 15 posts

Re: A jackal's nose for the job

#3
Wow, that is the interesting article, not the wikipedia one. Some excerpts:

"The groundbreaking approach is based on analyzing the bioelectric activity of the animals' brain and central nervous system"

"Because of errors in "signal behavior," the accuracy achieved by the traditional dog method is only 60 percent, says Zaripov. However, the new method being used has increased detection accuracy to nearly 100 percent."

Re: A jackal's nose for the job

#4

Wow, that is the interesting article, not the wikipedia one. Some excerpts: "The groundbreaking approach is based on analyzing the bioelectric activity of the animals' brain and central nervous system" "Because of errors in "signal behavior," the accuracy achieved by the traditional dog method is only 60 percent, says Zaripov. However, the new method being used has increased detection accuracy to nearly 100 percent."

Interesting that it also seems to handle false positives well:

> Among other things, the new method makes it possible to detect "a false sit" (when a dog sits to signal that it has found something). "In traditional signal behavior, 'a dog's trick' is hard to detect: The dog does not smile, it does not wink," says Zaripov. "In this case, however, an encephalogram shows that the dog has a conscience. Deep down it knows that it has made a mistake. The reaction of its central nervous system is completely different."

Re: A jackal's nose for the job

#5
Yes URL should be replaced with the rbth article.

So they developed a portable lie detector for dogs? And it works better than human lie detectors because the dogs don't understand it?

I wonder if a smart enough dog can figure out how to fool it to get a treat anyways.

Re: A jackal's nose for the job

#8
post #7
post #2

This article has a lot more information about the Sulimov dog: https://www.rbth.com/longreads/jackals/

Thanks! we've switched to that from https://en.wikipedia.org/wiki/Sulimov_dog .

It's nearing the 90 day mark. Are you going to prove yourself a liar by keeping me banned, and a coward for not letting anyone speak against you?

I don't mind either way. I'm just pointing out your inconsistencies.

Re: A jackal's nose for the job

#9
"It is not a breed," Sulimov explains. "We have reproduced the initial 'pre-breed norm' of a dog, the way it used to be 100,000 years ago. In other words before humans began to 'specialize' it, breeding it for hunting, guarding, etc. Any specialization limits possibilities."

I found this fascinating and the "Any specialization limits possibilities" reminds me of that Heinlein quote:

"A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects."

Re: A jackal's nose for the job

#10
post #7
post #2

This article has a lot more information about the Sulimov dog: https://www.rbth.com/longreads/jackals/

Thanks! we've switched to that from https://en.wikipedia.org/wiki/Sulimov_dog .

I think "High-tech jackal-dogs" would make a better submission title. It's used in the article as a section title and it seems like it better represents the content.
Post reply on HN