The ongoing Elon Twitter saga: "insert demographic" melts down

Status
Not open for further replies.
GordyR said:
I see, that's fair enough; although this feels a little bit like that interview now, don't you think?

Anyway, I'm a bit confused by other parts of your response and would like to pick you up on a couple of things:



This seems odd given that you were responding to someone who absolutely was talking specifically about the studies; and then in your response you said:



Why would you explicitly mention the data driven studies, even to the point of make a sarcastic comment about them not being real hard science, if you weren't talking about data driven studies?

I don't know, maybe it's me, but I don't think that it's unreasonable to think that someone who mentions data driven studies, is in fact talking about data driven studies.

Still, I appreciate the clarification.



Well no, not really.

In studies that set out to examine hate speech, you should expect to find hate speech explicitly defined, and if they've done their job properly then you should expect to find that definition properly adhered to throughout their research. If they weren't consistent in any way, then you should expect the peer review process to correct for their failings.



Perhaps reading the studies in question to find out precisely how they defined it would help?

Just a suggestion.
Click to expand...


Specifically, a range of terms were examined to see if frequency of use increased in correlation to the change in ownership and stated platform focus. The terms included vulgar and hostile terms for individuals based on race, religion, ethnicity, and orientation.
Click to expand...

So it's unclear exactly what words the study looked at and in what context the phrases were used. What is clear is that the number of times those words were used is statistically insignificant and infinitesimal compared to the amount of Tweets sent each day, we're talking around 500 Tweets per day out of 500 million. For the BBC journalist to suggest he's literally seen an increase in hate speech is laughable.
 
Roar87 said:
So it's unclear exactly what words the study looked at and in what context the phrases were used.
Click to expand...

I would consider this "study" to meet my request for an example that has failed to clearly define what they constitute to be hate speech. So congratulations.

If that really is the complete study and not merely a brief summary of their work, then it's a dreadful example of academic writing.

After a bit of digging it appears that they used a dataset of 517 terms compiled by Thomas Davidson, Dana Warmsley, Michael Macy, and Ingmar Weber. 2017. "Automated Hate Speech Detection and the Problem of Offensive Language." but I cannot absolutely confirm that without reaching out to them myself.

Here's the dataset if you're interested:

Hate Speech Identification - dataset by crowdflower

Tweets classified as hate speech, offensive language, or neither
data.world data.world

Apparently they then used Tweet binder to apply sentiment analysis in order to detect the context of the tweets that contained these phrases, and that's how they arose at their numbers.

Assuming the above is accurate then both the dataset and the methodology are likely to be fairly robust as this technique and dataset has been repeatedly used and re-validated in multiple peer reviewed papers over the last 4-5 years.

Regardless, even if their dataset was well defined and compiled, like the above; and even if their methodology was absolutely robust and their findings useful, the fact that they failed to explicitly mention exactly what phrases they were analysing is unforgiveable from an academic write-up perspective; at least in my opinion.

As such, you've absolutely met my request for an example - Although I do have my suspicions that the pdf you linked isn't actually the complete work.

But anyway, thankfully this paper appears to be alone in it's poor write-up, with far better written examples having also found similar trends:

Take this one from the ISD for instance, which found that antisemitic tweets more than doubled after Musk's acquisition:


There's more out there I came across in the 15 minutes i've dedicated to it, but I need to get back to work so you'll have to forgive me.


Roar87 said:
What is clear is that the number of times those words were used is statistically insignificant and infinitesimal compared to the amount of Tweets sent each day. we're talking around 500 Tweets per day out of 500 million.
Click to expand...

I cannot agree with this at all.

Firstly, the actual totals themselves are largely irrelevant. When you're evaluating the impact of policy change, it's the percentage change which is what's important; and from the little data i've looked at on the back of this exchange between you and I, it seems fairly damning so far.

Also, the idea that a negative trend change on the back of a policy change is only important if it impacts a lot of people, is a pretty ghastly one. Apply that logic to anything else and see how it holds up. For example, the amount of children who are sexually abused is infinitesimal relative to the amount of children who are not. If a government puts in a policy that results in pedarists and paedophiles feeling emboldened to commit more crimes, do we not evaluate the effect of that policy based upon its relative change, rather than on the actual totals itself?
 
Last edited:
BBC journalist James Clayton sent an email to the Twitter boss asking the "government-funded media" label. Usually, reporters receive a statement from the company's media team, but this time, Elon Musk replied personally. Clayton then seized the opportunity to request him an interview. And Musk responded with: "Let's do it tonight."
Click to expand...

BBC journo requests interview with Elon Musk, Elon Musk grants him an interview the same day.
So surely if anyone is prepared for this it is the person who requested the interview, yet there are people foaming at the mouth and spluttering out words whose meaning resemble that the journalist was unprepared and was ambushed by Elon Musk.

To those who feel the BBC journo was ambushed, do you often find yourself calling people you have not spoken to in years or have only heard of and then screaming down the phone "WHY ARE YOU AMBUSHING ME I AM UNPREPARED TO TALK TO YOU".
Surely when YOU make the call you have a conversation topic in mind so if anything the respondant is ill prepared for your topic of conversation...no ? my words and their meanings are alien to you?
 
Roar87 said:
Well, this New York Times article from December 2022 mentions that slurs against gay men and black Americans had increased, I'm not sure it's possible to know that it was a slur used against a Black American, or even if the slur was also being used by a Black American. How is it possible to identify the skin colour of someone being insulted and someone doing the insulting based on a blank Twitter profile? How can this possibly be considered science? The article mentions that slurs increased from 1,200 per day to
3,800 per day, yet one individual could set up a couple of bot accounts that would easily account for that increase. There are over 500 million Tweets sent each day, yet we're talking about an increase of 2,000 that could actually have been made by one user setting up a couple of bots, and this is evidence of a rise in hate speech on Twitter. The studies are laughably bad. Then you get some idiotic journalist suffering from confirmation bias posing it as a serious question to Elon who then rips him apart accordingly, probably because he took 2 minutes to look at the data himself and realised what a joke it was
Click to expand...

more recent article the beeb guy was probably trying to recall, with some attempts at experiments to show tweets the just for you feed could now be polluted with,

Musk has disabled (well commercialized) api to enable people to track hateful tweets, so it is pretty difficult to prove/disprove he has tidied it up, equally the open source exposure of the algorithm,
doesn't prove it won't radicalize people.

Musk's termininology during bbc interview of non-regretted twitter time, was interesting, he said, lower than tik-tok - that's cognitive dissonance ? (people don't like to acknowledge they wasted their time)
equally he had no viewpoint on the tik-tok proposed ban - don't want to alienate the chinese (manufacturers)

e: really the liar excerpt from the interview exemplifies the problem of out-of context quotes, radicalized/amplified, on twitter, to drive the conclusion that the whole interview was crap.
 
Last edited:
Amnesia said:
BBC journo requests interview with Elon Musk, Elon Musk grants him an interview the same day.
So surely if anyone is prepared for this it is the person who requested the interview, yet there are people foaming at the mouth and spluttering out words whose meaning resemble that the journalist was unprepared and was ambushed by Elon Musk.

To those who feel the BBC journo was ambushed, do you often find yourself calling people you have not spoken to in years or have only heard of and then screaming down the phone "WHY ARE YOU AMBUSHING ME I AM UNPREPARED TO TALK TO YOU".
Surely when YOU make the call you have a conversation topic in mind so if anything the respondant is ill prepared for your topic of conversation...no ? my words and their meanings are alien to you?
Click to expand...

:cry:

Prove you read the thread and aren't "frothing" by quoting this point of view and making sure 100% that it's the up to date freshest quote from that person. You have 3 pages to check, go go go.
 
JonRGV250 said:

Absolutely superb.
This should get the frothers going
Click to expand...

My favourite part is at 0:50 where the journalist says he "doesn't actually use that for you feed anymore because he doesn't particularly like it, and knows many people who are similar"

Hea perfectly described the media bubble - you must absolutely never expose yourself to opinions or views that aren't aligned with the great vision, comrade!
 
Hotwired said:
It's a good pressure sales pitch, he's in demand and he can demand an interview right on the spot.

They got played and Elon got to have the theatre that he prepared for.
Click to expand...

wesimmo said:
From the brief clip in that Tweet, it's more like debating someone who has been banned from SC for being unable to cite evidence for their claims.

The reporter was poorly prepared, he should have had the various data driven studies that have been done and proven a rise in 'hate speech' to hand.
Click to expand...

Hotwired said:
After a look around at news coverage, they don't really care. They wanted to hear Elons answers.

I see a few people who are not pleased the interviewer wasn't prepared enough to debate on the subjects being questioned and would rather they had the chance at the interview.

Give it 24h to see if anything changes in the coverage but seems like it'll go down as a wasted opportunity.
Click to expand...

All these are correct, but it still was a fail from Elon too. The guy gave Elon the opportunity to address the issue that users are reporting they are having. Rather than answer, Elon said "arr you having the issue" and then seemed to suggest that as the journalist wasn't having the issues as much as others were, the issue didn't exist. Imagine to be those users, listening to this and thinking great, Elon will give us an answer, and the answer is "it doesn't exist because it doesn't happen to 1 single user". It's almost as dumb of a reply as the users on here that say how great twitter is now because it runs so fast for them and everyone else reporting issues are lying.

Imagine this was Elon's reply.

Journo "can you address the recent reports in hate speech since you took over twitter"

Elon "I have heard that there are select users that feel their experience has worsened due to some of the changes we have made. Twitter has had some major changes recently, some will be for the worse but we hope over all most are for the better, and we continue to try to improve and move in the right direction. To those who feel hate speech has risen for them, I hope they will see an improvement over the coming months as we continue to find new ways to improve the experience, twitter blue being a prime example of how we hope this will reduce bots and hate across the board, pushing only quality content to your screens, and they stay with us as we work hard to try to get a good experience for all."

But Elon didn't do that did he as he's an idiot that instead would rather have the news about how a journo being a lier because he doesn't experience the same level as hate as other users do.

Do better Elon
 
Last edited:
These guys really don't want to waste the article they'd written.

In an NPR article about NPR it says they'll silence their 52 Twitter accounts because of the labelling, specifically the one Elon gave them after supposedly researching on Wikipedia: https://www.npr.org/2023/04/12/1169269161/npr-leaves-twitter-government-funded-media-label

Except in the BBC/Elon interview the same morning he was asked about the labels and said he would change it for the BBC and NPR to "publicly funded" instead of the somewhat suspect "government funded" label stuck next to every post they make.

Anyway, apart from that they do have some good points about:

1) Does leaving Twitter even matter to their reach (not clear that it would)
2) Elon Musk branding every tweet made by an account with labels he decides on
3) The splashback on reputation and maybe even operations if these labels are taken badly by the public.

Maybe it's not a wasted walk out though, if you think these are games of bad taste the only way to defend yourself is to walk off the platform. If you have to pander to the business owner you're corrupting yourself.
 
Hotwired said:
3) The splashback on reputation and maybe even operations if these labels are taken badly by the public.
Click to expand...
A point I'd not considered at all, but definitely a good one. I wonder if twitter are liable for damage to a companies reputation due to the false labelling.

And then is it worth it for a company to stick around on twitter and potentially run the risk that one day Elon decides to make a new rule that hurts their brand, for the little reward twitter brings.

Perhaps the solution would have been to give these accounts the option to pick from a predefined list of options, and then have them manually checked, before it then goes live in x weeks. Would have avoided a lot of this negative press Elon can't stop creating for himself
 
Last edited:
Poor BBC interviewer James Clayton thought this was his big day, had a list of lies ready to throw at an unprepared Elon Musk.
It didn't take long for poor James to throw in the towel though and unfortunately for James there was no referee to save him, this truly was a humiliating beating for James Clayton and all those Elon Musk haters.
James practically begged Elon Musk for mercy several times but Elon Musk did not relent, James begged as he knew all those who had hope in him must be in tears right now at the disappointment they were witnessing, how did Elon Musk manage to so easily dismantle the lies.
James was in shock, those who believed in him left in dismay, angry hot wired guys screaming, beating their mothers in anger blaming them for the failure of James Clayton.

Today will be remembered as how not to do an interview with Elon Musk when all you have is lies.

sMLOvD8.jpeg
Yes it was "so good".
 
Last edited:
Amnesia said:
Poor BBC interviewer James Clayton thought this was his big day, had a list of lies ready to throw at an unprepared Elon Musk.
It didn't take long for poor James to throw in the towel though and unfortunately for James there was no referee to save him, this truly was a humiliating beating for James Clayton and all those Elon Musk haters.
James practically begged Elon Musk for mercy several times but Elon Musk did not relent, James begged as he knew all those who had hope in him must be in tears right now at the disappointment they were witnessing, how did Elon Musk manage to so easily dismantle the lies.
James was in shock, those who believed in him left in dismay, angry hot wired guys screaming beating their mothers in anger blaming them for the failure of James Clayton.

Today will be remembered as to how not to do an interview with Elon Musk when all you have is lies.
sMLOvD8.jpeg

Yes it was "so good".
Click to expand...

I knew it, you read one post and started "foaming" as you put it, blind to anything else :cry:
 
Hotwired said:
:cry:
Click to expand...
Is this smiley used to mask your emotions?


See reading this post without the smiley it looks a lot more angry
Hotwired said:
I knew it, you read one post and started "foaming" as you put it, blind to anything else
Click to expand...
 
Amnesia said:
Is this smiley used to mask your emotions?


See reading this post without the smiley it looks a lot more angry
Click to expand...

It's the one for when someone starts typing the fanfic in their imagination.

It was half drivel the first time so I said prove it since I knew exactly what cherry picking you were up to, only 3 pages to find the quote backing it up.

Now it's pure drivel and you're offended by the laughing emoji while posting a GIGANTIC picture of a laughing Elon Musk.

Have another one.

:cry:
 
Status
Not open for further replies.
Back
Top Bottom