Artificial intelligence officially gets scary

[vimothy]

but its taking over anyway

 

[vimothy]

maybe this is bad

 

[luka]

its bad and its worse if its getting worse at it cos it means all i watch on youtube is pitbull rescue videos and i dont even like dogs or have any interest i just consume what im given without any attempt to intervene

 

[vimothy]

we're all gonna be in our little yt holes, except we're not gonna be watching content made by anyone - instead its intentionless garbage that ticks some statistical box which satisfies the underlying algorithm. that's the ultimate end state

 

[luka]

remember all that stuff about the creepy childrens videos? they might have been automatically generated i cant remember

 

[vimothy]

haha exactly that

 

[vimothy]

that's the future

 

[luka]

 

[vimothy]

the future is already here its just not distributed evenly

 

[version]

Eating Disorder Helpline Fires Staff, Transitions to Chatbot After Unionization

The chatbot is named "Tessa" and will replace the entire Helpline program starting June 1.

www.vice.com

 

[Clinamenic]

imagine an ai version of tiktok, where no one has created the video, pushing it in front of a human face, forever

TikTok is testing an AI chatbot called Tako

The bot appears capable of answering a wide range of queries.

www.theverge.com

 

[IdleRich]

its bad and its worse if its getting worse at it cos it means all i watch on youtube is pitbull rescue videos and i dont even like dogs or have any interest i just consume what im given without any attempt to intervene

The algorithm sees what you watch and then works out your taste so it can send you more stuff. If you don't like what it sends you then clearly your taste is somehow wrong or inconsistent. You need to sort that out.

With comedy, I've seen a few things on YouTube or elsewhere where either the comments or even the title says something like "AI could never write this!" which suggests to me that they are worried.

Certainly a lot of comedy is extremely formulaic, there are a number of joke structures that occur again and again. Even if there are some kinds of jokes that are beyond AI just now, so what? You could easily make a programme out of only the most formulaic jokes and ninety percent of people wouldn't complain.

 

[thirdform]

its bad and its worse if its getting worse at it cos it means all i watch on youtube is pitbull rescue videos and i dont even like dogs or have any interest i just consume what im given without any attempt to intervene

Don't you have a sapiosexual fetish? I'm sure @sus can help train the algorithms for you.

 

[wektor]

vimothy is sort of right here but I think the bigger problem is not whether they quality of AI generated content (whatever it may be) is poor so far.
it improves on weekly basis, eventually it will be good.
imo we will reach a point when chat gpt will be highly refined and extremely convincing/great at communication, yet will still present "facts" that have nothing to do with reality

 

[wektor]

maybe AI generated content is just another new ? same as TV and internet was.
it just takes a bit for the older generations to adapt. ten years ago it would be very difficult to convince your parents that some piece of information found on the internet is factual. nowadays they are the ones willing to take the word of any random schizoposter from a facebook group discussing tinfoil hat design.

 

[wektor]

same as with TV, it will take a bit of distance and plenty of domain-specific awareness/knowledge to distinguish between absolute bollocks and the somewhat real

 

[wektor]

re: @luka computers are already playing chess with each other, increasingly. I am fully convinced most social media will be 95% ads and spambots, 5% of the human content will be either people who stick to it out of habit and maybe some conviction (see: evening standard readers) or people more or less consciously choosing to do so while framing it as a transgressive new medium or whatever.

 

[wektor]

there already was a point a few years ago where most of the meme content on my facebook wall was generated and posted by bots, mostly set up as a vanity project by comp sci undergrad students. it was the best shit, sadly facebook hit them very hard and restricted the api usage.

that's the thing, ig and facebook spam/inappropriate content detection are already more likely to have a problem with you based on some easy to miss policies, while the russian botnet prompting you with a potential iphone win/shein coupon knows how to dance around the regulations all too well.

 

[luka]

re: @luka computers are already playing chess with each other, increasingly. I am fully convinced most social media will be 95% ads and spambots, 5% of the human content will be either people who stick to it out of habit and maybe some conviction (see: evening standard readers) or people more or less consciously choosing to do so while framing it as a transgressive new medium or whatever.

i wasn't suggesting computers cant play chess against one another. i was saying that they do and it doesn't capture peoples attention.

 

[wektor]

i wasn't suggesting computers cant play chess against one another. i was saying that they do and it doesn't capture peoples attention.

have you heard of virtual dog racing?