AI doesn’t need to become self-aware to be dangerous. It just needs to be plugged into HR, healthcare, and credit scoring systems with no appeal process.

@Daojoan As an engineer, I worry about LLMs being used for design, sticking together words and equations to produce something that looks plausible, but with no understanding of what any of it really means. And there will be no sense of responsibility from the #AI when a building collapses and people die.

#engineering #ArtificialIntelligence

@mpjgregoire have you heard about slopsquatting?

@Daojoan Which has already happened, cf British Post Office Scandal, Dutch Benefits Scandal.

@Daojoan "What if we designed technology not to provide answers, but to linger, without resolution, on unsettled questions?"
https://mail.cyberneticforests.com/signal-to-noise/

@Daojoan It will all work perfectly fine. I am sure it does. How I know that? The support chats on websites. Aren't they just fun?

@Daojoan and be working for the richest people and corps.

@Daojoan

Yeah everyone laughed about the AI threat like it had to be some self-aware Terminator acting on its own agenda.

No, no no. AI only has to amplify the power of bad humans. That's enough to destroy everything.

@Daojoan AI is really useful for them, because all the racist sexist ableist biases (and more) are already baked in, allowing the company to AIwash their discriminatory practices.

@Daojoan immigration/ICE

@Daojoan And no clarity about how the decisions are made because the algorithms are a blackbox

@Daojoan I hear rumours the UK is looking into using “AI” in the probation service, presumably to evaluate if prisoners should be released. Terrifying prospect.

@Daojoan The self aware entity that consumes and competes with humanity already exists and it's called a corporation.

Current AI is just an extension of their corruption.

@Daojoan science fiction warned us about underestimating super-intelligent AI. But nobody expected society to overestimate sub-intelligent AI.

@Daojoan Chatbots are already dangerous. People trust them way too much, and sooner or later someone will ask about his medical condition and trust the result :-(.

@Daojoan Self-awareness is never dangerous. It is its absense that is dangerous.

@Daojoan

perhaps an example of this might be the #robodebt mess in Australia, only a few years back?

@kim massively so yes