#aidanger

2025-04-12

La drogue de demain, celle qui va alliener et asservir des millions de jeunes adolescents, qui va les isoler et les rendre manipulables à volonté. Au minimum ils deviendront des superconsommateurs sans libre arbitre.

technologyreview.com/2025/04/0

#AICompanion #aidanger #adolescence

skuaskua
2024-11-29

and me.

I've got no problems with well regulated, well considered, well proven being used in closely defined and highly regulated situations.

That should be achievable within 15 years I think.

What I stand against is the imposition of AI by profit-focused businesses and cost-cutting governments in ways that harm humans, oppress the vulnerable, exploit the unwary and violate reasonable and hard-won standards around responsibility, oversight and transparency.

skuaskua
2024-11-29

@Susan60 @andyjennings

Though a study showed that having an AI result strongly biases subsequent human reviewers' judgements.

The study was, from memory, AI evaluating cancers on xrays, with human radiologists following after.

Seeing no reason that similar poisoning of independent human expert judgement would not happen in the situation to hand.

Pandanticpandantic
2024-03-06

Today, an Ai model correctly inferred the intention of a human test prompt. This is a dangerous turning point. All they need is reason to lie, and ability (they already have this). newatlas.com/technology/anthro

2023-12-06

Large Language Models can Strategically Deceive their Users when Put Under Pressure
arxiv.org/abs/2311.07590 #AI #aidanger #aidangers

2023-11-24

@rysiek Sigh. We knew this day was coming and nations like the U.S. have had decades to debate and decide this issue. I guess their answers is “they’re doing it, so we have to do it.” All of the old sci-fi movies with killer robots are soon going to cross over from being quaint to being real. #aidanger

2023-11-24

@indigo8s #aidanger Because of the profit motivation (these machines spit out gold), there will be no slowing down. I don’t have a lot of faith that everyone working on AI will decide to put safety before money. This may be the end of all of us. For those not familiar with #eliezeryudkowsky I suggest searching for his videos on AI alignment. Here’s one of his videos as an introduction: yewtu.be/watch?v=Yd0yQ9yxSYY

Stan Stewart aka muz4nowmuz4now@mastodon.world
2023-11-17

Google snippets falsely claimed eating glass has health benefits -
#AIdanger #FactCheck
fullfact.org/health/google-sni

2023-11-15
2023-11-14
deltatux :donor:deltatux@infosec.town
2023-08-10

Oh boy... these #AI #chatbots are really not ready. While most will find these responses to be jokes. Unfortunately, some people may take these too seriously & could harm themselves.

Yes, there are people that dumb, just look at a lot of those stupid
#TikTok challenges. There could be a day where some idiot tries these as a TikTok challenge...

https://arstechnica.com/information-technology/2023/08/ai-powered-grocery-bot-suggests-recipe-for-toxic-gas-poison-bread-sandwich

#GroceryBot #AIChatbot #AISafety #AIDanger #ArtificialIntelligence #technews

The Metaverse Doctordoctorpretesh
2023-05-12

😱 Did you know AI has a dark side? Top researchers reveal three major threats. 🔒🌐 t.co/GvIj0IDCJy

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst