Facebook Is Designing Its Own Chips to Help Filter Live Videos

Facebook Is Designing Its Own Chips to Help Filter Live Videos
May 25, 2018
https://www.bloomberg.com/news/arti...ning-its-own-chips-to-help-filter-live-videos

 

I wonder what percentage of Facebook videos are "live". I would think it would be relatively small such that, for now, they could be parsed out and screened before actually being posted for others to view.

 

Facebook doesn't want to pay for that. Also, it's psychologically hard on human reviewers.

 

And too many live vids to monitor...

 

I agree. But doing nothing is getting them into trouble. And for sure, developing, testing, modifying, testing, testing, testing your own "chips" as proposed will surely cost many $millions if not ultimately $billions. As I noted, I said "for now", they could maybe staff it.

Is there really? That's what I am wondering about. I think by far, most videos are recorded, then uploaded to FB at a later date. I really don't think there are that many (relatively speaking) that stream "live".

 

They are reviewing video. I recall an article not long ago, focusing on reviewer stress.

 

@Bill_Bright In the first quarter of 2017, 1 in 5 Facebook videos was a live one.

 

i dont know for USA , but in Asia (and especially the country i live in) live vids when you are in a club, party or karaoke is very common. Not saying hot chicks showing-off their beauty (which is quite pleasant).

 

"Facebook Is Designing Its Own Chips to Help Filter Live Videos"

I guarantee there will no nefarious abilities built into these chips. I promise.

 

Thanks Roger.

1 in 5 is only 20% so I was right that most, by far, are not live. But 1 in 5 is still a lot, and more than I expected.