@polanco Maybe the future of online is tricking the bots into talking with each other.

@polanco Maybe we could fool the marketers and astroturfers that dominate places like reddit into engaging with our own army of AI bots while we all just enjoy the internet somewhere else

@matty how can we enjoy _our_ internet (say, gemini) when its content could as well be AI generated?

@polanco Honestly I don't know. Some sort of william gibson esque active defense measures that run some kind of AI auto detection maybe? That probably won't be very practical

Really it might be necessary to just design new protocols that can't be used by bots. I wonder how that could be done.

Also, government regulation could be very effective against advertising / astroturfing. But that isn't very likely to happen, and if it does it won't have enough teeth.

@polanco I really am concerned about the future as AI undergoes commoditization. What is to stop a nefarious news outlet from just running a deep fake as fact? Could police create fake evidence with AI? What is to stop governments or special interest groups using bots to upvote or write supportive content astroturfing a position?

I have a feeling this will all start to happen with some alarming regularity. Hell people are already making deepfakes now for fun on youtube.

@polanco It's bad enough that popular news media manufactures consent so easily - I think AI will make their control absolute.

@matty I think the answer to that lies in the (political) eye of the beholder.

E.g. anarcho-capitalists could argue that once people notice X-news faked their information/pushed their own agenda, they'd lose popularity and eventually go bankrupt.

I personally think the State should somehow prevent, regulate and punish such behaviour, but as you said, what's gonna stop the goverment of today from twisting and breaking the rules in their own favor?

@polanco Also the media really is a tool of the class that runs the state. Noam Chomsky wrote about this in Manufacturing Consent. I see AI just as another tool in their toolbox. It's a digital extension of their continuous war on the non ruling class. Until the state actually works for us, I wouldn't hold my breath waiting for laws protecting us from artifical intelligence that works to serve the status quo.

@matty thank you for reminding me I have to check out Neuromancer.

Yeah, afaik, goverments should have _us_, their citizens, as their top priority/in their best interests, but sadly it seems that everyday we're treated more like cattle, and what's worse, people doesn't realise, or seem to be ok with it.

@polanco FWIW, GPT-3 without a human editor rambles a lot. Looking at its comments:


I would immediately lose attention if someone talks this much aimlessly.

@ashwinvis yeah, but what scares me the most is how it can be so consistent expressing an idea. About the rambling part, I hope it never learns how to be succinct.

Just the kind of reaction I'd expect from a bot 😅

@polanco @tregeagle @polanco
But you'd be wrong, I'm a dog, don't tell anyone.
Sign in to participate in the conversation
Mastodon @ SDF

"I appreciate SDF but it's a general-purpose server and the name doesn't make it obvious that it's about art." - Eugen Rochko