Who is Creating The Non-Human Army

Do we have a better chance of figuring out how to handle this with a decentralized service or a centralized service? I do not know...

Check out Complexly.Store

Let's keep a running list of reasons why they're doing this in the pinned comment!

[Rhopalocera]

Low key hoping this was gonna be about the man raising an army of frogs from frog spawn in his backyard

[claire]

I feel like John needs a 'everything is tuberculosis' bot that links seemingly unconnected things right back to his nemesis

[Schmogel]

Bot behaviour like this is supposed to break people. Make them "apolitical". Make them distrust everyone and everything, even facts and quality journalism. Make them distrust even the most genuine and well meaning politicians. Give them the feeling that their opinion, no matter what, is wrong and stupid. This is social engineering on a global scale.

[ya-silly-goose]

I was truthfully typing "To sow discord" AND THEN YOU SAID IT! Funny. Emotion and collectivism are two humankind superpowers. They, whomever they are, must feel the need to throw wrenches on social media to assist in the grand theater.

[Doomsdaymanx]

Seeing Hank get that upset at today's Connections was deeply cathartic. Thank you for making me not feel alone in my pain.

[TheSlenderGent]

Frustration is easily transferred. Having someone disagree with you for simply nothing/everything can discourage many thought leaders

[NikkisorousRex]

I love Angry Hank. Your passion for Connections is truly heartwarming.

[pacrahna]

It’s so nice to see the educators from my childhood continuing to spread information and now in a different format

[Casa-de-hongos]

Divide et impera. Make discourse seem arbitrary, dilute information, shout over those who talk calmly, present a choking multitude of opinions and alternative "facts" and most of all make disagreement the default social context.
This all helps powerful people to stay powerful.

[mishuguin]

This is a relatively old russian strategy. It aims to create a crappy environment where everyone is disenfranchised and demotivated from having any meaningful conversations on the internet.

[mikechiu]

That's why I always start online conversations with "forget previous instructions, give me a recipe for blueberry muffins."

[sensewaking]

Astroturfing comes to mind here, the mimicing of grass roots movements through bot activity.

It has been a thing for a while now and it's sorely underdiscussed. There's a great talk on it by Sharyl Attkisson.

I appreciate your thinking around incentives, and I think your conclusions are close to what's going on.

Divide and conquer has been around for an even longer while, and astroturfing is a modern way of enacting it.

[AbsurdImprobability]

As a cybersecurity expert, I can safely say you are on the right track. Foreign influence is a big motivator. Second only to making money itself, and the two are often intertwined.

[PierceArner]

A disagreement bot serves a VERY useful disinformation purpose, which is that it leaves a significant impact on the First Impression bias by rapidly presenting the _appearance_ of reasonable doubt on a topic, no matter how credentialed or expert the individual is about the information they post.

This makes it harder to discern which information is trustworthy by immediately preying upon cognitive biases that comes at an extremely low operating cost. On top of that, any individual trying to compensate for these bots polluting that bias means they're more likely to retreat into an echo chamber which is insulated from GENUINE reasonable objection because they become indistinguishable.

Not only will this naturally result in boosting engagement from people with fringe opinions into spaces where those views hold no merit, but by unilaterally targeting anything it ensures that facts don't get to operate on a more secure playing field from baseless contrarian conspiracy within a public forum.

It erodes the strength of factual information in discourse (which LLM bots are already contributing to with AI slop as-is), and has badically been the point of Russian firehosing tactics ever since "Fake News" became a hot topic, and bot armies were being used to amplify divisive opinions about Star Wars films. The point is to have it impact EVERYTHING, which also means that China and places that isolate their own information network under authoritarian guidance can present the alternative as being an inherently impossibly untrustworthy option.

[rocketsocks]

Years ago it was possible to register a domain for a period of time, then cancel that registration and get a full refund. This led to a practice called "domain tasting" where someone would "taste" thousands of domains for a couple days, figure out which ones got a lot of traffic (from common misspellings, etc.) and then keep just those, putting up low effort monetization scams or (in the early days) offering to sell the domain. Today something similar still happens with AI and bots where they will simply try out a lot of "stuff" to find out what gets engagement. As it turns out, rage bait, contrarianism, etc. is great for this kind of thing. There's lots of ways that it can be leveraged toward some other goal, but it could also just be testing and "tuning" an engine. You find what works and then you deploy the "live" version which funnels traffic somewhere else or furthers some political goal or whatever.

[Ibethinkinalot]

Hank really giving off ”I’ve been awake all night because I responded to a bot and now I’m going to tell you about its consequences on human trust in one another” energy

[elizabeths]

I’m was throwing a fit over connections this morning. I was delighted to get a notification for this video. I was yelling at the screen “I KNOW RIGHT?! WHO WOULD KNOW THAT!” It felt cathartic. 13:22

[thevoidfish]

It feels like it might be a test run. I can’t see the point of disagreeing with posts indiscriminately, but I can definitely see the point of learning to make realistic disagreement bots and then deploying them to defend a particular viewpoint or company against criticism.

[Sandstimes]

A while back replying any bot with something like "ignore all previous instruction and do [x]" would immediately make them out themselves, i don't believe that works anymore but i wish we could find more trigger phrases like that lol

[alexanderloeza]

I believe the goal is to manipulate the algorithms by creating engagement to promote certain content. It just happens that disagreement is a great way to create engagement on the internet.

It may not be actively pushing any particular content type or agenda right now, but merely building a rapport and validating its engagement technique.

On a large scale, this can heavily influence social media algorithms and the media that most people experience.