YouTube still has an ad problem

[ThinkingSapien]

YouTube still has an ad problem

Google had problems with this last year, as a result some advertisers pulled their ads. A significant portion of Google’s money is made from advertisements. Part of their reaction to this last year was to demonetize videos and pages that dealt with certain areas of discussions that the brands didn’t want to associate with and redefining what makes a video eligible for ads and the revenue from them.
Engadget:

YouTube’s problem with automated ads appearing on extremist channels isn’t getting any better. A recent CNN investigation revealed that companies including Adidas, Amazon, Hershey, Netflix and Under Armour – and even the US government – had their ads placed on unsavory channels.

As you’d expect, advertisers are incredibly upset and have vowed to leave the platform until changes are made. After an absence, Proctor & Gamble agreed to return to YouTube, but only on the condition that it can pick which channels and videos that its ads for Crest, Tide and other products appear. Bloomberg writes that the advertisements will run on “fewer than 10,000 channels.”

“We paused advertising, and for the past year, we’ve worked extensively with YouTube to improve brand safety,” a Procter & Gamble spokesperson told Bloomberg. “We now feel the right measures are in place for P&G brands to have the option to advertise on YouTube.”
[…]
It’s easy to see why YouTube’s algorithms matched the ads with the videos, but that’s not an excuse for it happening in the first place. A bulk of Google’s coffers are filled by automated ads; the company should have better safeguards in place as a result. Same goes for Facebook. Google loves touting its prowess in the AI field, but situations like this prove that a human touch is still necessary.

 

[SuperLuigi]

Thanks for sharing this.

Youtube has had 2 big problems:

1. Originally, they drug their feet on censoring porn, terrorism propaganda and dealing with copyright issues.
2. Because today they are focused on trying to silence conservatives and various anti-SJWs, they have less manpower to address REAL problems like some alarming instances of child abuse posted on their platform.
3. Some of these private companies have had absolutely horrible policies. Really, the business models used here are breathtakingly bad.

Nintendo FINALLY got it through its thick skull that it could profit from walk-throughs being posted, but unless you indicated in the contract, they would be entitled to half your earnings from a channel EVEN if most of the vids had nothing to do with them!

The clampdown on Star Trek fanfiction videos was just as bad, although not directly related to Youtube corporate. That was a decision made by CBS out of sheer fear that no one would watch their new series. A lot of Star Trek fan videos were very well done and had involved professional actors for 30 years. Fans are attracted to this because studio budgets sometimes limit the scope of the starship setting.

I think a lot of corporations are in denial about what is going on and are otherwise completely clueless as to who their customers are and what their preferences are. Namely that 95% of people aren’t going to boycott a company because an ad that hurts their feelings shows before a video. People with lives (and spending capital) don’t care about stuff like that unless it’s really out there like child abuse or ISIS propaganda.

 

[1Lord1Faith]

I think a lot of corporations are in denial about what is going on and are otherwise completely clueless as to who their customers are and what their preferences are.

I actually think AI is making their cluelessness worse. I find most examples of AI to be annoying rather than helpful. It’s probably not meant to be helpful anyway, at least not to us. It’s really just a tool for gathering information about people. Information which corporations can’t seem to use, even for their own benefit. They just sell it to someone else, and tell them how useful it is.

 

[ThinkingSapien]

I think a lot of corporations are in denial about what is going on and are otherwise completely clueless as to who their customers are and what their preferences are. Namely that 95% of people aren’t going to boycott a company because an ad that hurts their feelings shows before a video.

I think their customers are primarily the advertisers; the exception being viewers that pay to not see advertisements (YouTubeRed).

I actually think AI is making their cluelessness worse. I find most examples of AI to be annoying rather than helpful. It’s probably not meant to be helpful anyway, a

Interesting to see what new categories that the Facebook AI came up with for advertisements to target.

 

[SuperLuigi]

AI is not the magic machines that we see in Sci-Fi. There’s so many issues with machine learning. It’s like an automated phone system on steroids.

 

[FirstFiveEighth]

That is what the robots want you to think.

 

[ThinkingSapien]

AI is not the magic machines that we see in Sci-Fi.

What we see in SciFi might largely be described as artificial consciousness. Real life AI might be better described as a form of computational analysis and pattern detection. One of the difficulties in working with some types of AI is that the people responsible for it’s creation don’t necessarily know how it works. With other types of computing a person may have thought out some bit of logic and expressed it in code form. With AI (especially neural networks) a program is allowed to make adjustments to the set the setting for each artifical “neuron” on its own. While it is possible to probe the individual neurons to get their settings the understanding of the entire system, simple though the foundations may be, is too complex for human understanding. Outcomes that are not the intentions of the creators is possible.