Threads vs Twitter: misinformation vs uoᴉʇɐɯɹoɟuᴉsᴉɯ

[Isaac]

Backstory: In case you're not aware, there's a new social echo-chamber app called Threads, created by Instagram, which means it's by Facebook and that fucker Mark Suckaburg. It's their attempt to compete against Twitter, a dumpster fire, porn-infested, bot hell-hole, which has gone from bad to worse since Elon Muskymusk has been squeezing profitability out of his overpriced message board. You should use neither. That being said, I've been using Twitter since Muskymusk took over, and Threads since I was allowed to, because I want to experience these apps firsthand.

This has been entertaining for me and here's why:

• Twitter is actively blocking all positive news about Threads on its platform. My proof? Twitter is ensuring related tags are all about privacy issues and memes against Threads. This is abnormal, imo. People love complaining about Twitter, so why is is absent on Twitter? Comparable to this is what's happened on Reddit, where Lemmy is promoted on just about every subreddit. Reddit has restrained themselves against banning speech about Reddit alternatives, though they have in the past. Also ★■◆● /u/spez. I think Twitter would have the same amount of pro-Threads posts if Musk wasn't cracking down on them.

• Threads is blocking anything they don't like, including my posts against the Chinese Communist Party (CCP): I still need more proof of this. But I think my posts on this topic are getting censored. If this is true, it's a massive ★■◆●ing deal and Zuck can go ★■◆● himself.

• The account that was tracking Muskymusk’s plane, which was banned on Twitter after Musk took ownership, moved over to Threads: Lmao.

• Threads was warning users against following profiles with conservative views: This was recently removed, but it exposes their bias and sends the message that Threads is actively working against anything it sees as misinformation, which is most likely most conservative talking points. It’s also entertaining because Facebook has never given a ★■◆● about misinformation from anything other than American right-wing political misinformation.

• I enjoy the absence of bots on Threads right now and it’s great: So on Twitter, there are so many bot accounts. I should know, I’ve coded 3 bots for Twitter that don’t use the API. It’s stupid easy. Why? Because Twitter’s web interface has many points where a scraper can mountaineer and spelunk around the HTML and act like a regular user session. Threads does not have any web interface or any API/dev documentation I can find. Maybe there’s an API I can't find, but my spider senses on Threads does not detect any bots. Everyone feels like real people joking around. Creating a bot that mimics the official app is beyond my ability, and I'm guessing many other people are going to have trouble too. But all the NFT spammers are probably pissed at how they can’t spam their scams like they do on Twitter through their 500 fake accounts. This was an eye-opener for me. Maybe web browser access is something that should go away?

Not really news, but would love to hear about any other juicy tidbits you might have heard about this internet drama.

edit: I'm editing this still

edit: I'm done. If anything is unclear, let me know.

[Darth Wang]

I'm just glad I never use social media.

[Isaac]

That's good, though the dbb is absolutely a form of social media. We even have a flat earther and a possible Nazi. We even had a Russian operative in 2015. Not to mention the gas lighters and those who like to be lit. We're basically a scaled down Twitter.

[Tunnelcat]

Yeah, I'm bad enough here. Sorry guys. I don't need Fakebook, Twine or Titter to ruin my day. Besides, why give any more money and prestige to a couple of stuck up rich pricks who don't have anything better to do with their money than allow people, and bots, to en masse to insult or put down each other? You must be a gluten for punishment Isaac to even be on either platform. And, are we sure Sigma wasn't a bot?

[Isaac]

And, are we sure Sigma wasn't a bot?

That was a mean joke. I apologize for saying that.

[Tunnelcat]

Don't apologize. He was just a mouthpiece for his master, Vladimir Putin.

[Isaac]

Don't apologize. He was just a mouthpiece for his master, Vladimir Putin.

I'd love to ask him how he's dogged the draft.

[Tunnelcat]

Maybe he hasn't. He could be an older coder/programmer working in Putin's cyber spook department. A lot of the younger Russians don't like Putin or his little war and refuse to listen to his state propaganda like the old farts that watch TV tend to do. Speaking of hackers, I've just had to freeze all my credit at all three credit agencies. There were data breaches at the Oregon DMV where they got names, SS numbers and birth dates of millions of Oregonians. Another at a financial institution that also has my information, all because they both were using the same information sharing software that was compromised, this time by Chinese hackers I believe.

[TigerRaptor]

The only thing I like about Facebook is the groups, and finding out about upcoming products.

Like this.
image hosting

[Isaac]

Is that an electric riding mower?

[Tunnelcat]

Have you seen the price? A kings ransom at $4,499.99. Then there's the batteries that cost an arm and a leg to replace when they go bad.

[Krom]

What I really want from ego is for them to partner with shark and make an upright vacuum.

[Isaac]

I want atomic batteries already. Caution be damned. I want the motor to burn out before I run out of power.

[woodchip]

Isaac, You seem to be pretty knowledgeable, what do you think AI and deep fakes will have on social media

[Isaac]

Isaac, You seem to be pretty knowledgeable, what do you think AI and deep fakes will have on social media

This is all my opinion but...

I think we've seen nothing yet and all the biggest problems aren't going to be from corporate run ai platforms, but the open source home-brew communities.

It's laughably uncompetitive at the moment, but take a look at the home-brew competition to chatgpt:

https://huggingface.co/models?pipeline_ ... generation

Note that some need 60+GB of VRAM to run using CUDA


There's a huge problem with ai art producing child pornography. There's no controlling it. The people who make CP using ai are able to train models using existing models as a base. The same is going to happen to large language models, aka the locally hosted home-brew versions of chatgpt.

In 3 years, you'll be able to run a unrestricted large language models that will help you do anything. "My neighbor has a 2020 ford focus. How can I steal it and not get caught". We won't get to this point until consumer grade GPUs provide 80GB of VRAM or more.

Now to answer your question, once you run these LLMs on your local device with no restrictions, you can run complex phishing scams, where your LLM pretends to be an entire company full of different people and do everything from setting up a website, buying a website, setting up an email server. It will be able to make phone calls and make use of locally hosted realistic voice models to sound like different people. Sounds far fetched? It's allready kind of possible using chatgpt's limited api: Now imagine the capabilities with no limits on your desktop PC running the equivalent of an A100.
On social media you'd ask it to make 100 twitter accounts and generate profiles. It could then pretend to be these people for a month to gain followers, you suddenly have a powerful weapon in the social media space. Chatgpt is limited by how long a comment history could be. Your locally run version might have a comment history hundreds of times longer depending on your VRAM size.

I give it 3 years, but if I'm wrong we'll be in this ai hell in 5 for sure.