Ah that algorithms! The discussion is not so much about humans vs machines, but about #explainability and #consistency of the decision making. You won’t justify decisions by data alone, but in terms of first principles, while taking rebuttals into account.
Algorithms are much less biased than humans, but you'd never know from the current brouhaha about them:https://hbr.org/2018/07/want-less-biased-decisions-use-algorithms …
-
-
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Algorithms are as biased as the humans who implement or use them. Just like humans, not all algorithms are the same.
-
I'd say algorithms can be biased in a very different way than humans. Take the learning algorithm that outputs the same decision all the time. Very different bias. I think the point is that we need to think about the sources of "bias" and even the "right" definition of it.
Keskustelun loppu
Uusi keskustelu -
-
-
no doubt that algorithms could be less biased. yet, there is no compelling reason why increased accuracy should result in more favourable outcomes. logically, reduced biases will lead to reduced access to goods/services for certain groups of people (whilst favouring others).
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Disagree, algorithms are trained on data which captures human decision making. So generally they are just about as biased as humans.
-
Maybe yes, maybe no. I'd say more often, algorithms are trained on data that is easy to collect, regardless of whether human decisions are involved! So they can be biased more, or less than humans. Not really black and white. Does this make sense?
- Näytä vastaukset
Uusi keskustelu -
-
-
There's a whole firld of study dedicated to the trade off between bias and variance on estimators, and how model constraints cause bias. Algorithms are inherently biased, which is why they shouldn't be cited without an expert view on their limitations.
-
You’re grossly misunderstanding the meaning of ‘bias’ in statistics.
- Näytä vastaukset
Uusi keskustelu -
-
-
That's a gross generalization. As gross as "algorithms are bad", which BTW, is not the hypothesis behind
@FrankPasquale,@PopTechWorks or@mathbabedotorg 's publications.Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
But who writes the algorithms?
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.