writing program administration | writing assessment | language & literacy ideologies | trans rhetorics
#LLM science summaries were nearly five times more likely than human-authored science summaries to contain broad generalizations (95% CI [3.06, 7.70], p < 0.001).
And newer models over-generalized more; not less!
doi.org/10.1098/rsos...
#philSci
spitfirenews.com/p/grok-csam-...
spitfirenews.com/p/grok-csam-...
Yes, there are some unexpected outputs. But if you'd chatbot creates sexual images of young children, either the programmers don't care or THEY EXPLICITLY WANTED IT TO DO THAT.
This is bullshit. A chatbot is not something that can apologize.
Pretending otherwise is simple laundering these companies' bullshit about what AI is, while diffusing blame away from the human beings that developed and released this system.
Yes, there are some unexpected outputs. But if you'd chatbot creates sexual images of young children, either the programmers don't care or THEY EXPLICITLY WANTED IT TO DO THAT.
www.reuters.com/legal/litiga...
www.reuters.com/legal/litiga...
tidal just put it at the very top of my most listened December playlist
www.youtube.com/watch?v=VjRo...
tidal just put it at the very top of my most listened December playlist
www.youtube.com/watch?v=VjRo...
It was this googledoc:
It was this googledoc: