In AI, those who can't do research call themselves ethicists and tell researchers what to do.
Näytä tämä ketju
-
Hey, this is awful.
-
-
Hey, this just confirms that it's awful.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Name-dropping Kahneman in a screed about how being forced to examine one's work for biases is harmful is about as text-book bias blind spot as I think I've *ever* seen. I suggest you have not used your own system 2 thought to consider your work as thoroughly as you could.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
‘Mathematically incapable of bias’ is a staggering nonsense, Pedro. They’re not programmed by some Newtonian variant of the Platonic ideal, they’re programmed by people and there are multiple real-works examples to show how flawed the results can be.
-
* real world. My phone’s AI once again making a poor guess about my poor typing because it was programmed by someone who thought frequency was a better bet for programming word choice than grammar and context.
Keskustelun loppu
Uusi keskustelu -
-
-
"We must stop politicizing artificial intelligence", writes a man in a politically conservative magazine, calling those advocating for ethical use of AI 'liberals' and 'progressives'
-
"Data can have biases, of course, as can data scientists." So.... Since you agree this bias exists, how do you propose to check for, or account for, biases in the data used to train models (or biases in the people who decide on what metrics to measure model performance, etc)?
- Näytä vastaukset
Uusi keskustelu -
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.