News for hackers. Stuff that matters.
Re:Perspective: Google API that detects toxic comments by kristianc on Saturday August 12, 2017 @05:29PM (#15000082)
An API for censorship, that really is something.

One of our great facets as humans is the ability to sort through information and come to opinions based on it.

We come pre-equipped with our own API for critical reasoning, honed by thousands of years of evolution, and with a training set of everything we've ever done and seen.

Why would I want to use a copy of that, that at best will only replicate human biases?

And that risks sidelining perfectly valid viewpoints? Does something have to be expressed in college-educated grammar to even be read with this in place? What are the consequences of that?

My first instinct, and I hope yours, was to push back the slider.

EDIT: I pasted the first paragraph of the Google diversity memo into it, and it got a "10% likely to be perceived as toxic" score. Make of that what you will.

Re:Perspective: Google API that detects toxic comments by exabrial on Saturday August 12, 2017 @05:50PM (#15000205)
I was thinking about doing the same thing actually with the memo. The whole thing is such a shame, the timing with this api is no doubt strange
Re:Perspective: Google API that detects toxic comments by dvfjsdhgfv on Saturday August 12, 2017 @06:01PM (#15000259)
That's definitely a management problem. Someone thought since the product is ready we should announce it. They're not stupid and understand very well what the current climate is, but decided to announce this peculiar tool anyway. Well, they'll definitely get quite some feedback in the next days.
Re:Perspective: Google API that detects toxic comments by kristianc on Saturday August 12, 2017 @06:03PM (#15000266)
The API for not releasing products into a media shitstorm is clearly still under development...
Re:Perspective: Google API that detects toxic comments by ocdtrekkie on Saturday August 12, 2017 @06:32PM (#15000410)
As a friendly warning, I suspect one word in your comment might be perceived as toxic.
Re:Perspective: Google API that detects toxic comments by chocolateboy on Saturday August 12, 2017 @06:34PM (#15000418)
> They're not stupid and understand very well what the current climate is, but decided to announce this peculiar tool anyway.

It's not new: https://news.ycombinator.com/item?id=13713443

Re:Perspective: Google API that detects toxic comments by dvfjsdhgfv on Saturday August 12, 2017 @06:37PM (#15000430)
Thanks for pointing this out, the timing seemed really strange.
Re:Perspective: Google API that detects toxic comments by DanBC on Saturday August 12, 2017 @06:43PM (#15000445)
> We come pre-equipped with our own API for critical reasoning,

We clearly unambiguously don't. People suck at many critical reasoning tasks. Look how poor people are at assessing risk for the most obvious example.

Re:Perspective: Google API that detects toxic comments by maehwasu on Saturday August 12, 2017 @09:18PM (#15000993)
Yes, that's why it's a good thing I have Google to make sure I'm doing my struggle session correctly!
Re:Perspective: Google API that detects toxic comments by frankosaurus on Saturday August 12, 2017 @06:41PM (#15000437)
> We come pre-equipped with our own API for critical reasoning, honed by thousands of years of evolution, and with a training set of everything we've ever done and seen.

> Why would I want to use a copy of that, that at best will only replicate human biases?

Scale. Who has time to read thousands of trollish youtube comments to find the handful of insightful gems?

Re:Perspective: Google API that detects toxic comments by kristianc on Saturday August 12, 2017 @06:45PM (#15000452)
Luckily this is a problem that the market has solved for you, as YouTube has an upvoting mechanism.

A similar upvoting mechanism is the reason why're you're reading this story as opposed to others.