Update: Following my tweet on Thursday, Google fixed its autocomplete promoting that "civil war is inevitable"

Here's what Google autocomplete for "civil war is " looked like before Thursday:

THREAD Image
Today, there is only one suggestion saying that civil war is coming: 2/ Image
But Google didn't fix it for other languages, for instance for French there is the same thing: 3/ Image
In Spanish we also have one suggestion that that states that civil war is inevitable: 4/ Image
Why does Google suggest that? Does the algorithm know something we don't? Having worked at Google, I don't think so. I think fearmongering is efficient for clicks, so Google *promotes* fearmongering.

5/
Now, imagine what many people think when they see "civil war is inevitable" coming from Google. Many must think there has be some truth to it. They might think they need to take action.

6/
Conclusion: when exposing dangerous algorithmic bugs, Google takes action.

Finding them one by one is not enough. So I'll continue working with researchers to create tools to help anyone find those bugs.

7/
PS: Here's the link to my Tweet from Thursday:

8/

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Guillaume Chaslot

Guillaume Chaslot Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @gchaslot

7 Dec 20
YouTube's deep learning AI boosted massively two news channels in the US last month:

➡️Newsmax TV went from 3M to 110M views/month (x33!)
➡️New Tang Dynasty TV (NTD): 6M to 80M views/m (x13)

They both made wild claims of voter fraud. Why does the AI like them?

Thread ⬇️
I already made a thread about Newsmax TV last week (see below), so let's focus on NTD

2/
New Tang Dynasty TV was funded by the Falun Gong religious movement, that got banned in China, and thousands of followers have died under Chinese custody.

This channel existed for years, but its views just exploded last month. Why?

3/
en.wikipedia.org/wiki/Falun_Gong
Read 10 tweets
4 Nov 19
Is YouTube's algo still taking people down the rabbit hole?

Yes.

Right now, the algorithm is promoting to millions of teens this conspiracy that literally says:

"We ask you to suspend your disbelief and take a journey down the rabbit hole" (1:37)


1/5
It's on a channel called "After Skool", so clearly targeting kids/teens.

It was so massively recommended that it reached 1 million views in few days

2/5
One of the top comments is revealing:

"Hey YouTube? What's this doing in my recommendations? If I didn't have a basic understanding of science and the ability to recognize logical fallacies this could've sent me down the path to conspiracy paranoia"

3/5
Read 5 tweets
1 Nov 19
Thread

I discovered that a Chinese anti-American conspiracy theory was promoted by the millions by YouTube's AI

@OliviaGoldhill reported on it, and @YouTube reacted to our findings

1/
Right after publication of the article, @YouTube added a warning on the Chinese video that it's "inappropriate for some users"

For who?

NBA officials😂?



2/
Why hide this video? If anything, people should be able to discover that this is the official Chinese propaganda line.

Why did this video get millions and millions of free ads? Was it truthful, or just efficient at gaming the up-next algorithm? 🧐

3/
Read 13 tweets
4 Oct 19
In the current context, which YouTube channel is particularly promoted by YouTube's Artificial Intelligence?

A Ukrainian channel that copy-pastes Fox News videos

Read that again, but slowly. 1/

algotransparency.org/?date=03-10-20…
That channel did >2M views in the last few days. It was an increase of 3,080.6% month-to-month.

It's now called "BREAKING NEWS"

Full stats:
socialblade.com/youtube/channe…

2/
It was called "Florans UA" before the 30th of September.

Note that the channel describes itself as Ukrainian, but we can't be certain it's Ukrainian. In any case, it's performing exceptionally well.

3/
Read 7 tweets
14 Jul 19
Thread

My first op-ed in @WIRED: how the AI feedback loops I helped build at YouTube can amplify our worst inclinations, and what to do about it.

wired.com/story/the-toxi…

1/
Earlier this year a YouTuber showed how YouTube's recommendation algorithm was pushing thousands of users towards sexually suggestive videos of children, used by a network of pedophiles.

YouTube bans sexual videos. What happened?

2/
At YouTube, we designed the AI for engagement. Hence, if pedophiles spend more time on YouTube than other users, the job of the AI will become to try to *increase* their numbers.

3/
Read 24 tweets
22 May 19
.@coffeebreak_YT shows how the trending tab is heavily biased towards TV channels, spoiling native YouTube creators

Let's try to investigate why this happens. 1/
Having worked at YouTube, I know they don't intentionally bias algorithms, but they can pick a metric and not investigate enough to see if it creates bias.

Here's my theory about it. 2/
In order to show something in trending, they probably check that the video doesn't offend people. This seems reasonable. 3/
Read 7 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!