|
|
In the Name of God بسم الله
Science/Health/Economics
Discuss science, technology, health, medicine, fitness, economics, finance, business, and engineering.
Note: Forum is not intended as replacement for professional advice.
3,869 topics in this forum
-
- 550 replies
- 24.4k views
-
- 5 replies
- 125 views
-
- 13 replies
- 199 views
-
- 714 replies
- 30.1k views
-
- 387 replies
- 12.7k views
-
- 1 reply
- 65 views
-
- 0 replies
- 51 views
-
- 1 reply
- 76 views
-
- 620 replies
- 24.7k views
-
- 39 replies
- 732 views
-
- 3 replies
- 100 views
-
- 2.4k replies
- 108.7k views
-
- 0 replies
- 70 views
-
- 41 replies
- 1.6k views
-
- 1 reply
- 134 views
-
- 0 replies
- 147 views
-
- 1 reply
- 179 views
-
- 12 replies
- 1.1k views
-
- 2 replies
- 118 views
-
- 0 replies
- 136 views
-
- 0 replies
- 184 views
-
- 0 replies
- 97 views
-
- 2 replies
- 190 views
-
- 1 reply
- 320 views
-
- 0 replies
- 77 views
-
- 1 reply
- 276 views
-
- 213 replies
- 21.3k views
-
- 1 reply
- 620 views
-
- 5 replies
- 325 views
-
- 6 replies
- 286 views
-
Recently Browsing 0 members
- No registered users viewing this page.
-
Upcoming Events
No upcoming events found -
Posts
-
The bots today are not like those of, say the 00s or the 10s. These bots are very, very sophisticated. Part of my job, in my real work, is training AI (artificial intelligence) models, not for bots, but the process is the same. I know the technology from the backend. I'm still learning it, but I have more information than most people have. So without going into too much detail and boring people, AI is only as sophisticated as the data you give it. AI relies on gigantic volumes of data (lots and lots of examples) in order to train the AI model. For example, with text posts, the AI model is fed, given examples of, millions of posts on a given subject. When it is trained, the 'trainor' can give preference to certain types of posts, score them higher so that the AI bot will try to emulate them. This is called assigning weights. It's one step in the training process. The main criteria, for weights, is popularity. They will always try to emulate the popular posts as closely as possible, because the goal of the spam bot is always to get many people to see their posts, and popular posts spread and are given automatic credibility. They can weigh posts heavier that talk about Ahl Al Bayt((عليه السلام)), Rasoulallah(p.b.u.h), etc, if they are talking to Muslims and Shia. Then the bot will incorporate those references into it's posts. They can also feed in other posts and weigh them. For example if they are talking to Iranians, the millions of examples would be in Farsi instead of English and they could weigh posts that talk about Iranian Nationalism or women's rights higher so that the bot brings up these more in it's responses. After the training process is done, and the bot is unleashed, the average person, even a sophisticated average person, would have a hard time figuring out which posts are from bots and which are from real people, because most of the content of the posts is just normal stuff (my name is __, I live in ___, I like eating chocolate cake, I love my mom, Paris is nice in the spring, etc). That's why it sounds like a normal person, until it delivers it's 'payload', which is the content that it is engineered to echo and propagate. I use the term payload because it is engineered exactly like a military weapon. To get to the target (which is the psyche and sentiments of a certain type of person) and then deliver it's munition, which is falsehood which is in someone's interest to propagate. The other thing bots will often do is 'hijack' an account. They will crack the password of a real person's account (pretty easy to do since most people make weak passwords, don't use 2 factor authentication, and social media companies are very lax when it comes to security, by design) and then use that person's account to post so it seems like the posts are coming from a real person. These are called 'zombie accounts', and there are millions of them. This is the new world we are living in. There are bots posting right now on SC. When we figure them out, we ban them, but even we (even us tech people) have a hard time figuring out which are the bots, just from their posts. -
-
How the unvaxxed can protect themselves This dude properly understands his target market. -
This is absolutely true. Under a capitalist system if you're not paying, you are the product.
-
I gotta be honest, I have no idea what y'all are talking about. I've got and I use social media but all I ever see is people I actually know on Facebook or Instagram. I don't use Snapchat or tiktok. My Twitter is a little crazy, but I really don't think most of who I see there are bots - there is no way a bot could be so chaotic and full of personality as most of the strangers I read. Definitely I'm getting a filtered view, we all are, but I don't know why I'm not seeing all the evil that so many other people are talking about.
-