the API is called perspective (http://perspectiveapi.com/)

http://www.forbes.com/sites/kalevleetaru/2017/02/23/fighting-words-not-ideas-googles-new-ai-powered-toxic-speech-filter-is-the-right-approach/#6dc28a99712b

I really like online conversations, like the ones that take place on Okayplayer because they are civil. There are times I want to jump in on discussions on other social media platforms but don't because the commentors just seem to be vile or petty or crass. Anything that cuts down on cyberbullying or harassment is alright by me.

The free speech concerns are valid to a degree. Ideas shouldn't be censored and what is hurtful to one group is acceptable in another. But there are so many pockets of the Internet where trolling is out of control and if there is no means of policing speech by humans, than an algorithm is the next best thing. Algorithms are only as good as those programming them, but I can't see that this new technology could be worse than the current methods of policing online content.

Over time, this technology will likely become more effective at determining fair and decent comments from crude or hurtful or hateful speech.

i wish this could be a simple issue of free speech good, censorship bad. i do see the point, however.

i'm a big believer in (and earnestly try to be a practitioner of) civility, but i believe that civility should come from within, both, to be hokey, inside one's heart and by the example set by others. i believe in self-policing. you should put your best self forward, and if you act up then others within the community should clearly convey that is not ok. it shouldn't come from a robot.

every community is different. as you know, there are places online where trolling and what we would both probably qualify as hate speech and hateful rhetoric are allowed to run wild. i think part of being a mature adult is knowing not to put up with nonsense. if it's entirely within an online realm and is clearly acceptable to the community, it makes more sense to part ways than expect change. i've spent a lot of time online from BBSs, AOL, and IRC to forums, comment sections, and social media and have the hard earned wisdom to show for it. i know my tolerance level and when something is a lost cause. i like to think i've also learned how better to treat people in my online communications based on a lot of trial and error. it's too easy to let a script decide that. that said, the internet is not the U.S. free speech is what we have, or strive for, here. it's not a right online. communication is a privilege. there are many instances where there are minors involved and where the so-called real world and online intersect, and there should be safeguards to prevent bullying and harassment. i can see the utility of this tool for that. i just think blocking and ignoring abusive users is a more mature way for adults dealing with the problem than resorting to censorship. censorship is still better than banning as the first and only recourse, as often happens. i also suspect those intent on being abusive will just find ways around the script.

relying on google to set the tone of discussion all over the web strikes me as similar to americans wanting every foreign land to have the comforts of home for them (mcdonald's, hard rock cafe, etc.). God forbid you put forth the effort to immerse yourself in someone else's (online sub)culture and not expect everyone else to cater to your personal sensibilities. the homogenization of the internet is almost as troublesome to me as the idea of censorship. censorship should be a last recourse, not a first line of defense, but ultimately up to the standards set by the particular community and the goals of each owner and operator. there's room for everyone and everything. we aren't owed a place to remain undisturbed at every table. if there are communities where such a script is a helpful nudge toward someone capable of expressing his or her ideas more civilly, great, but ideally that should have been learned by example. a world and an internet where AI restricts what we say and how we say it is a lot less rich and interesting to me and should be a red flag to any firm believer in open dialog and an open internet. the lover of language in me dies a little with the forceful restriction of how we can communicate with each other. individuality needs to be embraced now maybe more than ever.

i really do sympathize with those who want their communities to be more civil and think this is the solution, i just wish it didn't come to this. i guess that speaks to the state of humanity.

I think the entire internet should be like 4chan where you can post whatever the hell you want. The issue with internet speech is that it's anonymous. This API attempts to quell the the symptom of that rather than fix a root cause.

IMO, the two root causes of the "problem" of internet speech is that 1. internet speech is anonymous and therefore mostly free of consequence, and 2. people take it seriously. I would much rather see internet speech become unregulated and unmoderated (other than blatant spam/ads) to the point where people stop taking it as seriously. Internet speech is no more valid than speech scribbled onto the wall of a shitter in Sharpie. Could you have an intelligent, thoughtful, and productive conversation on the wall of a shitter? Certainly. But you can and do also get "CALL CARLA FOR A GOOD TIME 555-555-5555." I would much rather see the policing of internet speech fall on the individual who is viewing it rather than entrust it to giant corporations who already hold all of the cards.

Viewing anonymous internet speech is a choice. Every person who makes that choice should be responsible for their own feelings.

>the API is called perspective (http://perspectiveapi.com/)
>
>http://www.forbes.com/sites/kalevleetaru/2017/02/23/fighting-words-not-ideas-googles-new-ai-powered-toxic-speech-filter-is-the-right-approach/#6dc28a99712b

http://www.reuters.com/article/us-twitter-abuse-idUSKBN1684L4

there is so much junk on twitter so again i can sympathize with the decision, and their goals seem to be different from and maybe more pressing than this perspective API

So that is to be considered when they throw around this word "intelligent". AI most of the time imo is probably no more than a parlor trick and not real intelligence.

So that seems to be a bad idea. What they "invented" is a robot that censors people without a consciousness or paying a person to look at it with real eyes and a brain.

Porn detection software at photo labs works surprisingly O.K. so i dunno maybe it will work some of the time. The dick detection software malfunctions and gives a false positive a lot though because of all the phallic shapes around us.
peace
Key

http://keymusic.bandcamp.com/album/apophenia

https://keymusic.bandcamp.com/

retail photo development or something more forensic?

anyway, back to the topic at hand, i agree, it is censorship by robots, so the results will be a mixed bag at best