@GriftGod_ was targeted and successfully demonetized. He too, made videos about Alyssa.
He has to wait 90 days to re-apply.
The same people celebrating.
(con't)
Target: SmashJT
Status: Under Attac
@SmashJT is their current target. Not only is Alsyss suing him for alleged "harassment", Airbagged's army of groupies has him in the crosshairs using the same techniques for cancellation.
They are trying to get him banned from YouTube, or at least demonetized. They are combing through his videos as I post this.
(con't)
This is a revenge tour.
This is the cancel culture crowd.
They lost the battle last year and this is a petty move using every underhanded trick in the book to "get one" on the people who were most effective.
Going after people's sponsors, trying to cancel their conventions, their careers or more, is disgusting.
This is much bigger and far worse than what they did to Kabrutus. They've sunk to new lows.
Don't let it stand. Support your favorite creators and tell @YouTube about this TOS violating brigading going on.
This is not organic. This is coordinated, brigaded and organized through private Discord servers.
And they are not done yet. They found a loophole in the Youtube TOS that is so vague that it bascially lets them cancel anybody.
And they are using this, like a hammer, to go after everyone.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
There are massive complaints of deboosting of conservative voices in the new Nov-Dec X algo and I think I know why.
The Left has weaponized it.
- The new X algo is primed for signals to determine who gets shown on the feed.
- One of these signals is blocks. blocks are a huge negative signal.
- The Left for years, weaponized automated block lists. We warned X about them before. The Right doesn't use these much but the Left does en masse. The biggest conservative accounts are mass flagged on dozens if not hundreds of these lists.
- Regular X lists are also weaponized by the Left. The Left will create a ton of lists such as "Nazi" and worse and mass add conservative accounts. The Right does not do this. There is no way to mass remove yourself from these lists: you have to block the list creator one by one. I'm on probably 100-200 of these list which are created for a purpose.
- The purpose of these lists is to easily scan and notify Leftist activists when you post (and their bots). The X list doesn't filter and therefore not subject to the algo. This means activists looking for your posts will see them before your followers do. And there are thousands of activists doing this daily, monitoring many conservative accounts.
- The reason your followers don't see your posts right away anymore (have you noticed the lag between posting and views) is because the X Algo is scanning and testing your post first for engagement and sentiment before going wider with exposure. Since followers are the LAST to see vs the haters using those lists to scan and reply-guy to you, the HATER replies get posted FIRST. I believe this triggers the negative sentiment indicator, adding to deboosting the visibility of the post. You'll notice this in your own posts...hater go first...follower trickle in after.
None of this would be an issue if the Left played fair. But they don't. They weaponized block lists and X Lists long ago, and now that is being taken into account (I believe) more and more by the X algo.
Combined with early sentiment analysis and you have a possible reason why so many conservatives feel throttled today.
my recommendations:
- Find a way to combat automated block lists.
- Give users a way to mass remove themselves from X lists.
- Ignore early "hater" sentiment that is designed to get the algo to think your post is "negative" and de-prioritize you to followers and the For You tab.
On block lists, this used to use the Twitter API. Many of these lists are old and its probably less used now because of the cost of the API. But older conservative accounts from before the takeover are on these lists, and many Leftists used them. So we still have the lingering effects.
The media doesn't often lie directly. Instead they use a variety of techniques to maintain plausibility as purveyors of "fact."
1) Lie by image manipulation. This is subtle but effective. They color tinted Trump to be more orange, and they color-tinted Joe Rogan to look more sickly as they talked about him taking "horse pills." Another example is when they took pictures of the "noose" in Bubba Wallace's garage, which looked huge and head sized in the closely cropped photos, but turned out to be a 3 inch wide loop used to pull open the garage door.
A 🧵 (con't).
2) Lie by Omission. This is the most common one. They state a fact, but leave out another fact that gives important context. An example might be "The old woman was shot by police" but then neglect to mention the woman was a psychotic who was gunning her car directly at a policeman with intent to kill. This was done by Kotaku when reporting on Kabrutus while omitting the fact that SBI was trying to cancel him and take away his Steam library.
(con't)
3) Lie by False Authority. News articles will often quote "a source close to the issue" or "experts agree" without ever citing the source. This is sometimes necessary but it is nearly always abused. "Scientists agree" is also used very often.
WuChang's patch 1.3-1.4 performance increases are sus as hell. YouTuber Daniel Owens claims that all they did was REMOVE native resolution and FORCE you into lowered internal resolution WITHOUT telling you.
His video link and the results from his video in this thread. 🧵
WuChang is getting bad reviews on Steam because of performance issues. In response, the publisher/dev released patch 1.3 and 1.4 to address the issue.
The patch notes claim "supersampling resolution limits have been adjusted on select GPU models to prevent unintended performance degradation."
This is pretty vague, but more importantly, the game setting screen still reports native resolution at the highest settings.
HOWEVER, these gains are ONLY at the highest setting, and selecting lower resolution upscaling doesn't improve much at all.
Since scaling is on the GPU, any real performance increases should affect all resolution upscaling settings, improving them across the board.
But here they do not. You can see a shot of Daniel's excel sheet here with the numbers.
So what happened here? How could performance only increase on native resolution but not any of the other options? It should be across the board. If it's faster in native, it should be faster in all the other settings too.
Daniel's explanation is that they are lying to you. He claims the game NO LONGER RENDERS at native resolution. Instead, you are secretly forced into a lower internal resolution without your knowledge.
Worse yet, even if you had a bad ass GPU, you no longer have the ability to run it at native rez, according to Owens.
Dragon Maid has the most cited botched localization with the famous "Pesky Patriarchy" line. So I tried to fix it with @FishAudio AI.
Here is the original woke localization compared to the version I created by cloning the original Japanese voices and having them speak English.
How did I do?
I'm sponsored by Fish Audio but I can say what I like and will go over the pros and cons and techniques in this thread.
First, the original cringe woke version, my AI version in thread.
🧵(thread con't)
Here is the AI version.
I collected clips from the characters in original Japanese, which cloned it speaking Japanese. But Fish Audio will let you use the voice and change the language, so I used the English subtitles to replace the Japanese and got them to speak English.
It took about 10-15 seconds of audio for each character to get a clone.
How did it come out? Could this be used as a tool for fan localizations?
It wasn't that simple, more (con't)
First off, all the sounds were maid with AI, including the laugh and hesitation noises.
Fish audio lets you add tags like (hesitant) and (laugh), which I used here to create the sounds as well as the dialog.
To get the voices more consistent is the hard part. The AI will generate variations of the voice, often in different pitches. So what I did was lower the "Top P" setting to have it stick closer to the original, which seemed to work.
But even then, you are gonna get variations and I found it a bit challenging to keep the tone through multiple lines of dialog, but there are some workarounds for that here (con't)
Here are the 5 times Bungie has had to apologize for stealing art for their games. 🧵
Bungie is having a hard time. Marathon art theft is confirmed and they are trying to make it right.
It can happen, contractors or rogue artists on a deadline can cross the line under pressure.
But when the same problem happens 5 times in a row, it might be time to admit there is a deeper problem at the studio and gamers and artists patience is rightfully stretched thin.
(con't)
In 2021, Bungie had to apologize for using fan art of Xiva in a trailer for Destiny 2: The Witch Queen.
Original art by @relay314
(con't)
@relay314 In 2023, Bungie had to apologize for a vendor using fan art in Destiny 2's Lightfall cutscene.
The 1st weekend results are in, and Clair Obscur Expedition 33 has pretty much demolished AAA with a peak of 121,422.
This puts AAA games in the entire industry to shame and shows that something is very very wrong in the industry.
They key takeaways for me are (con't)🧵
AAA game teams are vastly bloated with hundreds of devs and dismissing returns for team size.
The entire Expedition team was 30 people.
The larger the team, the higher overhead in communication and co-ordination costs, and you also lose VISION as devs are just cogs by that point and passion plummets. (con't)
You do not need to have China labor costs to be competitive.
What you need are smaller teams who are very talented and skilled.
The rumor is that the game was VERY cost efficient. I'm hearing some crazy numbers like just 5.5M to make.
I suspect the number is closer to 15-30M, but this would still be astounding compared to the 300-500M AAA games are spending these days.